Search

Hemoglobin conjugated with poly(ethylene glycol) (PEG) acts as an oxygen carrier free in plasma, substituting red blood cells in supplementing oxygen in hypo-oxygenation pathologies. Given the complexity of oxygen delivery controls, subtle structural and functional differences in PEGylated hemoglobins might be associated with distinct physiological responses and, potentially, adverse effects. We have compared hemoglobin PEGylated under anaerobic conditions, called PEG-Hb(deoxy), with hemoglobin PEGylated under aerobic conditions, called PEG-Hb(oxy), a product that mimics Hemospan, produced by Sangart, Inc. SDS PAGE and MALDI-TOF analyses demonstrated that PEG conjugation yields products characterized by a broad distribution of PEG/hemoglobin ratios. The elution profiles in size-exclusion chromatography indicate that both products exhibit a more homogeneous distribution of molecular weight/hydrodynamic volume under deoxy conditions and at higher concentrations. PEG-Hb(oxy) shows high oxygen affinity, low modulation of allosteric effectors, almost no cooperativity, a fast and monophasic CO binding, and a limited dependence of functional properties on concentration, whereas PEG-Hb(deoxy) exhibits oxygen binding curves that significantly depend on protein concentration, and a slow CO binding, similar to native hemoglobin. PEGylated CO-hemoglobins, probed by flash photolysis, exhibited a lower amplitude for the geminate rebinding phase with respect to native hemoglobin and a negligible T state bimolecular CO rebinding phase. These findings are explained by an increased dissociation of PEGylated hemoglobins into dimers and perturbed T and R states with decreased quaternary transition rates. These features are more pronounced for PEG-Hb(oxy) than PEG-Hb(deoxy). The detected heterogeneity might be a source of adverse effects when PEGylated Hbs are used as blood substitutes.

Activation of the NO/cGMP pathway modulates smooth muscle cells relaxation and hence vasoconstriction, a major hindrance for the use of cell-free haemoglobin (Hb) as blood substitute, despite conjugation with 5-kDa maleimide poly(ethylene)-glycol (PEG) reduces vasoconstriction in vivo. We aimed at assessing how a recently developed PEGylated-Hb (Deoxy-PEGHb) and manipulation of the NO/cGMP pathway enable modulation of vasoconstriction in isolated rat hearts. Hearts were Langendorff-perfused with oxygenated Krebs-Henseleit (15 ml/min) while monitoring the coronary pressure (CPP) after injection (1 min) of 50 nM norepinephrine followed by a 1 microM Hb or Deoxy-PEGHb bolus, without altering the flow. Deoxy-PEGHb induced less vasoconstriction than Hb. Although the presence of PEG could contribute to vasoconstriction, Deoxy-PEGHb did not contain appreciable amounts of free PEG. Whereas reducing endothelial NO release by 0.2 mM L-NAME increased vasoconstriction, abolishing NO scavenging by Hb using its cyanomet derivative almost completely blunted it. Furthermore, maintaining intracellular cyclic GMP by inhibiting phosphodiesterase-5 with 0.02 mM sildenafil enabled control of Hb-induced vasoconstriction. We conclude that, although PEG-Hb represents a possible approach to limit Hb-induced vasoconstriction, manipulating the NO/cGMP pathway may provide a powerful way to circumvent this problem.

The 2-iminothiolane reaction with protein amino groups adds a spacer arm ending with a thiol group, which can be further treated with molecules carrying a maleimido ring. This approach is currently used for the preparation of a candidate 'blood substitute' in which human Hb (haemoglobin) is conjugated with long chains of PEG [poly(ethylene glycol)]. To identify the thiolation sites by MS, we have carried out the reaction using deoxyHb bound to inositol hexaphosphate to protect some of the residues crucial for function and NEM (N-ethylmaleimide) to block and stabilize the thiol groups prior to enzymatic digestion by trypsin and pepsin. Under the conditions for the attachment of 5-8 PEG chains per tetramer, the thiolated residues were Lys7, Lys11, Lys16, Lys56 and Lys139 and, with lower accessibility, Lys90, Lys99 and Lys60 of the a-chain and Lys8, Lys17, Lys59, Lys61 and Lys66 and, with lower accessibility, Lys65, Lys95 and Lys144 of the b-chain. The a-amino groups of a- and b-chains were not modified and the reaction of the Cysb93 residues with NEM was minor or absent. After the modification with thiolane and NEM of up to five to eight lysine residues per tetramer, the products retained a large proportion of the properties of native Hb, such as low oxygen affinity, co-operativity, effect of the modulators and stability to autoxidation. Under identical anaerobic conditions, the conjugation of the thiolated Hb tetramer with five or six chains of the maleimido derivative of 6 kDa PEG yielded products with diminished co-operativity, Hill coefficient h=1.3-1.5, still retaining a significant proportion of the effects of the modulators of oxygen affinity and stability to autoxidation. Co-operativity was apparently independent of the topological distribution of the PEGylated sites as obtained by treating partly the thiolated protein with NEM prior to PEGylation [poly(ethylene glycol)ation].

Allosteric proteins, such as hemoglobin, are assemblies of functional units, which undergo quaternary structural transitions in response to concentration changes of a specific ligand. Functional properties of hemoglobin ligation intermediates indicate that the tertiary structural changes induced by the ligand do not promote an equilibrium of quaternary structures.

Understanding mechanisms in cooperative proteins requires the analysis of the intermediate ligation states. The release of hydrogen ions at the intermediate states of native and chemically modified hemoglobin, known as the Bohr effect, is an indicator of the protein tertiary/quaternary transitions, useful for testing models of cooperativity. The Bohr effects due to ligation of one subunit of a dimer and two subunits across the dimer interface are not additive. The reductions of the Bohr effect due to the chemical modification of a Bohr group of one and two alpha or beta subunits are additive. The Bohr effects of monoliganded chemically modified hemoglobins indicate the additivity of the effects of ligation and chemical modification with the possible exception of ligation and chemical modification of the alpha subunits. These observations suggest that ligation of a subunit brings about a tertiary structure change of hemoglobin in the T quaternary structure, which breaks some salt bridges, releases hydrogen ions, and is signaled across the dimer interface in such a way that ligation of a second subunit in the adjacent dimer promotes the switch from the T to the R quaternary structure. The rupture of the salt bridges per se does not drive the transition.

Hemoglobin is a regulatory component of the oxygen transport to the tissues, and for decades has been a prototype to develop new strategies for the study of the structure/function relationships in proteins. One of the most difficult, and so far, unattained objectives of hemoglobin research has been the study of the hemoglobin molecules in a state of partial ligation with oxygen, or intermediates, as a means of testing theories of cooperativity. A cryogenic technique has been developed for the isolation, identification and quantification of the reaction intermediates of hemoglobin and CO, which in many aspects is a close approximation to the physiological ligand. The technical features that are crucial for the evaluation of the significance of the experimental data obtained using this technique and various approaches to the analysis of the data are reported. The discussion points out the importance of accessing direct information on the nature and concentrations of the intermediates in solution to clarify mechanisms of cooperativity as opposed to the less informative studies of the bulk properties of the solution.

Direct experimental resolution of the ligation intermediates for the reaction of human hemoglobin with CO reveals the distribution of ligated states as a function of saturation. At low saturation, binding of CO occurs with slightly higher affinity to the beta chains, but pairwise interactions are more pronounced between the alpha chains. At high saturation, the two chains tend to behave identically. The sequence of CO ligation reconstructed from the distribution of intermediates shows that the overall increase in CO affinity is 588-fold, but it is not distributed uniformly among the ligation steps. The affinity increases 16.5-fold in the second ligation step, 4.6-fold in the third ligation step, and 7.7-fold in the fourth ligation step. This pattern and the detailed distribution of ligated states cannot be immediately reconciled with the predictions of either the concerted allosteric model of Monod-Wyman-Changeux or the sequential model of Koshland-Nemethy-Filmer and underscore a more subtle mechanism for hemoglobin cooperativity.

A cryogenic technique for the isolation of the ligation intermediates in the association reaction between hemoglobin and carbon monoxide at 20 degrees C [Perrella, M., Davids, N., and Rossi-Bernardi, L. (1992) J. Biol. Chem. 267, 8744-8751] was used to study the effects of proton and chloride concentrations on the rates of the stepwise reactions. The reaction rate was observed to increase continuously in the course of the ligation process, yet the acceleration of the reaction after the binding of two ligand molecules, observed previously in 100 mM KCl, pH 7, was not observed at other pH values. At pH 6.3, such an acceleration occurred after the binding of three ligands, and at pH 8.5, a large acceleration was observed after the binding of the first ligand molecule. Greater CO binding to the beta chains was observed under all conditions, as in the previous study. The functional heterogeneity of the chains in the first ligation step increased with pH. The chloride concentration did not influence the distribution of the ligand between the alpha and beta chains at pH 6.3 and 8.5. At pH 7, less binding to the alpha chains was observed at 7 mM chloride with respect to 100 mM. The nature of the biliganded component isolated at pH 7 in 100 mM KCl and unresolved by the cryogenic technique was studied using a combination of cryogenic and noncryogenic isoelectric focusing. This component was a mixture of intermediates (alpha beta) (alpha CO beta CO), about 65%, and (alpha beta CO) (alpha CO beta), about 35%. The experimental data were compared with the distributions of intermediates calculated according to the Monod kinetic model assuming rapid and concerted transitions between two quaternary structures at each ligation step. The model provided a qualitative fit of the observed distributions of intermediates at acidic and neutral pH. A large discrepancy between the experimental observations and the predictions of the model was found at alkaline pH. The mechanism of the association reaction is discussed in the light of the available information on the tertiary/quaternary structures of the intermediates, as obtained from the studies of the deoxy/cyanomet model of ligation.

Free energy contributions to cooperativity by the eight ligation intermediates of human hemoglobin (Hb) have been characterized extensively using six oxygenation analogs [cf. Huang et al. (1996) Biophys. J. 71, 2094-2105, Table 2]. These unprecedented data bses have strongly supported the molecular code mechanism of Hb cooperativity [Ackers et al. (1992) science 255, 54-83]. The present study addresses a recent argument against this work [Shibayama et al. (1997) Biochemistry 36, 4375-4381] based on "free energy" determinations for a doubly-ligated species of the CN-met analog. Shibayama et al. (1997) have claimed that, in the hybridization experiments that have been used to determine free energy of the asymmetric "species[21]" tetramer, a portion of the bound cyanide is allegedly released from CN-met Hb during the incubation with deoxy Hb that is used to achieve hybrid equilibrium. These authors have claimed that cyanide release has resulted in extensive electron exchange between heme sites of the hybridizing sample, leading to incorrect evaluation of the equilibrium species population by the cryogenic techniques that have been employed. In this report, we demonstrate that neither appreciable cyanide loss nor electron exchange occurs with the methods that have been used extensively by our two laboratories for these equilibrium determinations [Perrella et al. (1990) Biophys. Chem. 35, 97-103; Daugherty et al. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 1110-1114]. An alternative experiment, which Shibayama et al. (1997) have carried out to illustrate their claim, does not evaluate a thermodynamic equilibrium property of the species [21] hybrid. The relevance of their newly-estimated "free energy" is therefore unclear. Nevertheless, Shibayama et al. (1997) have claimed that their proposed "free energy" (which is approximately 1.3 kcal more positive than the free energy of -11.4 kcal found independently by our two laboratories) renders invalid the molecular code mechanism of hemoglobin cooperativity. This representation is utterly without foundation since a free energy even more positive than suggested by Shibayama et al. (1997) would be fully consistent with the molecular code mechanism.

A technique is described for the rapid inactivation and removal of excess ferricyanide used for the non-cryogenic oxidation of the unliganded subunits of the intermediates in the association reaction between hemoglobin and carbon monoxide. Under these conditions the asymmetric oxidized intermediates, which dissociate into non-identical dimers, disproportionate into their parent tetramers and four species, Hb+, HbCO, alpha 2+ beta 2CO, alpha 2CO beta 2+, are isolated by non-cryogenic isoelectric focusing. The relative concentrations of species alpha 2CO beta 2+ and alpha 2+ beta 2CO measure the overall distribution of the ligand between the alpha and beta subunits in the association reaction. At 20 degrees C in 0.1 M KCl, pH 7, preferential CO binding to the beta subunits was observed, in agreement with observations made by the cryogenic technique for the isolation of the intermediates [M. Perrella, N. Davids and L. Rossi-Bernardi, J. Biol. Chem. 267 (1992) 8744].

The Bohr protons released by oxygen exposure of the unliganded subunits of intermediates (alpha +CN-beta) (alpha +CN-beta) and (alpha beta +CN-) (alpha beta +CN-) were obtained by titrations of concentrated solutions of these species. The Bohr protons released by oxygen exposure of the other intermediates were obtained from titrations of equilibrium mixtures of two parental species, (alpha beta) (alpha beta), (alpha +CN-beta) (alpha +CN-beta), (alpha beta +CN-) (alpha beta +CN-), and (alpha +CN-beta +CN-) (alpha +CN-beta +CN-), in which the concentration of the hybrid intermediate was determined by cryogenic electrophoretic techniques. The Bohr effect of the intermediates was calculated by subtracting the Bohr protons released by oxygen exposure of the intermediates from the total Bohr protons of deoxyhemoglobin at the same pH. The Bohr effects of intermediates (alpha +CN-beta) (alpha beta) and (alpha beta +CN-) (alpha beta) were similar and vanished at pH 8 where the total Bohr effect of deoxyhemoglobin is still significant. This suggests that the Bohr effect in these intermediates is tertiary in the quaternary T structure. The curve of the Bohr effect of intermediate (alpha +CN-beta +CN-) (alpha beta), which was close to the curve obtained by adding the Bohr effects of the two monoliganded intermediates at acidic and physiological pH values, was significantly different from the curve obtained by adding the Bohr effects of one liganded subunit of intermediate (alpha +CN-beta) (alpha +CN-beta) and one liganded subunit of intermediate (alpha beta +CN-) (alpha beta +CN-).(ABSTRACT TRUNCATED AT 250 WORDS)

The intermediates in the oxidation of deoxyhemoglobin by ferricyanide in 0.1 M KCl, at 20 degrees C and three pH values, were studied by cryogenic techniques. Data analysis was carried out according to a simple four rate constant model, ignoring the functional heterogeneity of the subunits, to simulate the time courses of the oxidation reaction, as studied by the stopped-flow technique [Antonini et al., (1965) Biochemistry 4, 345], which show anticooperativity at neutral pH and cooperativity at alkaline pH. Data analysis according to a 12 rate constant model indicated that the rate of oxidation of the beta subunit in the first oxidation reaction was 4 times faster than the rate of oxidation of the alpha subunit at pH 6.2 and 12 times faster at pH 8.5. The reactions involving the alpha subunit were noncooperative except for the last oxidation step at acid and neutral pH, but were cooperative at alkaline pH. The reactions involving the beta subunit were partly noncooperative and partly anticooperative. These complex mechanistic patterns suggest that a simple two-state model requiring the concerted transition of the tertiary structures of the subunits from the T to the R conformation is not adequate to interpret the oxidation reaction and that tertiary structures contribute, positively and negatively, to cooperativity. A structural hypothesis is suggested to explain the difference in the reactivities of the alpha and beta subunits.

The concentrations of the intermediates in the association reaction between human hemoglobin and CO at 20 degrees C, pH 7, under conditions of negligible dissociation of the ligand, were measured by cryogenic techniques. The monoligated species were predominant at all values of overall ligand bound studied. The analysis of the experimental data assuming a scheme of four consecutive reactions indicated that the binding rates increased in a continuous fashion. A significant acceleration after the binding of the second molecule of ligand occurred in the presence of 0.1 M KCl, but not with the addition of an excess of inositol hexaphosphate, indicating that major functional, and possibly structural, transitions occur at the diligated state. Differences in the concentrations of the intermediates in the same state of ligation were observed under all conditions. The analyses of the data on the basis of schemes of multiple pathways of reaction indicated that the beta subunits reacted about 1.5 times faster than the alpha subunits in the first ligation reaction. After the addition of inositol hexaphosphate, the alpha subunits reacted about 1.5 times faster than the beta subunits in the first ligation step, but the overall rate of the first CO binding step was unchanged.

The binding of carbon dioxide to human hemoglobin cross-linked between Lys alpha 99 residues with bis(3,5-di-bromosalicyl) fumarate was measured using manometric techniques. The binding of CO2 to unmodified hemoglobin can be described by two classes of sites with high and low affinities corresponding to the amino-terminal valines of the beta and alpha chains, respectively (Perrella, M., Kilmartin, J. V., Fogg, J., and Rossi-Bernardi, L. (1975b) Nature 256, 759-761. The cross-linked hemoglobin bound less CO2 than native hemoglobin at all CO2 concentrations in deoxygenated and liganded conformations, and the ligand-linked effect was reduced. Fitting the data to models of CO2 binding suggests that only half of the expected saturation with CO2 is possible. The remaining binding is described by a single affinity constant that for cross-linked deoxyhemoglobin is about two-thirds of the high affinity constant for deoxyhemoglobin A and that for cross-linked cyanomethemoglobin is equal to the high affinity constant for unmodified cyanomethemoglobin A or carbonmonoxyhemoglobin A. The low affinity binding constant for cross-linked hemoglobin in both the deoxygenated and liganded conformations is close to zero, which is significantly less than the affinity constants for either subunit binding site in unmodified hemoglobin. Comparing the low affinity sites in this modified hemoglobin to native hemoglobin suggests that cross-linking hemoglobin between Lys alpha 99 residues prevents CO2 binding at the alpha-subunit NH2 termini.

Crocodilian hemoglobin has a high intrinsic oxygen affinity but does not react with those organic phosphate esters that normally control the oxygen affinity of blood in higher vertebrates. Instead, its oxygen affinity is greatly lowered by CO2. The present study was undertaken to determine the nature of the CO2 binding to the hemoglobin of a crocodilian species, the Caiman, both qualitatively and quantitatively. The following parameters were measured: (a) carbamino compounds of deoxy- and oxyhemoglobin, (b) the effect of CO2 (at constant pH) on the oxygen affinity of Caiman hemoglobin, (c) total CO2 concentration of hemoglobin solutions at different pH and pCO2 values, and (d) the effect of CO2 on CD spectra of Caiman aquomethemoglobin. An analysis of the results of these measurements revealed that CO2 binding in the form of carbamate was not oxygen-linked and cannot, therefore, mediate the CO2 effect on the oxygen affinity. It was found, however, that 2 mol of bicarbonate can be bound/hemoglobin tetramer and that the association constant of the bicarbonate anion greatly depends upon the state of ligation. At pH 7.02 and 25 degrees C, a numerical value of 2.0 X 10(3) M-1 was obtained for deoxyhemoglobin, while for oxyhemoglobin no significant bicarbonate binding could be observed. At more alkaline pH (pH greater than or equal to 7.5), the association constant for deoxyhemoglobin decreases. Circular dichroism of Caiman aquomethemoglobin decreased considerably in the 287-nm region upon addition of CO2 at constant pH, an effect very similar to the one caused by inositol hexaphosphate in human aquomethemoglobin.

A human hemoglobin solution partially saturated with carbon monoxide was rapidly quenched at -25 degrees C into a hydro-organic buffer containing ferricyanide. Under the experimental conditions of pH, ionic strength, and buffer composition used in this work, it was found that the deoxy hemes were rapidly transformed into their met form, whereas practically no carbon monoxide-bound hemes were oxidized before the separation of the mixture from the oxidizing agent. As a preliminary step to the analysis of the resulting solution, carbonylhemoglobin solutions partially oxidized with ferricyanide were studied by isoelectric focusing at -25 degrees C under identical conditions. The relative position in the gel of all nine possible valence hybrids was established as follows (going from the anodic to the cathodic side of the gel) alpha CO2 beta CO2, (alpha CO beta +)(alpha CO beta CO) (alpha CO beta CO), (alpha CO2 beta +2), (alpha + beta CO), (alpha + beta +)-(alpha CO beta CO), (alpha + beta +)(alpha CO beta +), (alpha +2 beta CO2), (alpha + beta +)(alpha + beta CO), alpha +2 beta +2. When carbonylhemoglobin and methemoglobin were mixed in equal proportion at -25 degrees C and then analyzed by isoelectric focusing at the same temperature, it was found that the contribution of valence hybrids other than alpha CO2 beta CO2 and alpha +2 beta +2 to the total amount of hemoglobin in the gel was no more than 6%. When carbonylhemoglobin and deoxyhemoglobin were mixed in the same proportion and incubated at 20 degrees C so to allow the redistribution of the carbon monoxide molecules between all possible binding sites to occur, a substantially higher amount of valence hybrids, derived from the oxidation of intermediate compounds of hemoglobin with carbon monoxide, was found. The isoelectric focusing separation indicated the presence in the original solution of intermediate species other than carbonylhemoglobin and deoxyhemoglobin at a concentration of about 10% of the total.

The procedure of Perrella et al. (Perrella, M., Benazzi, L., Cremonesi, L., Vesely, S., Viggiano, G., and Rossi-Bernardi, L. (1983) J. Biol. Chem. 258, 4511-4517) for trapping the intermediate compounds between human hemoglobin and carbon monoxide was validated by quantitatively determining during the approach to equilibrium all the species present in a solution containing large amounts of intermediates. An accurate estimate of the intermediate compounds at 50% carbon monoxide saturation in 0.1 M KCl, pH 7, at 22 degrees C, allowed the calculation, according to Adair's scheme, of the four equilibrium constants. At 50% ligand saturation, the pool of intermediate species was about 12% of the total. A slightly greater concentration of tri-liganded than mono-liganded species was found. Carbon monoxide bound to beta chains in slightly greater excess with respect to alpha chains in both the mono- and tri-liganded species. The symmetrical bi-liganded intermediates, alpha 2 beta CO2 and alpha 2CO beta 2, were absent. The nature of the bi-liganded intermediate found to be present in detectable amounts by our technique has yet to be clarified: it could be either the asymmetrical species (alpha beta) (alpha CO beta CO) and (alpha beta CO) (alpha CO beta) or both of them. Such a finding on the functional heterogeneity among the four possible bi-liganded intermediates is consistent with hypotheses of the existence of more than two quaternary structures in the course of ligand binding to hemoglobin.

CO2-dissociation curves of concentrated human deoxy- and carbonmonoxyhemoglobin at 37 degrees, pH 7.6 to 7.0, PCO2 equal to 10 to 160 mm Hg, have been obtained by a rapid mixing and ion exchange technique. The CO2-dissociation curves for deoxyhemogloblin can only be fitted by assuming two classes of binding sites for carbon dioxide. The simplest way to account for the experimental data is to assume that the alpha-amino groups of the alpha and beta chains react with carbon dioxide with affinities that differ by at least a factor of 3. No difference in reactivity with CO2 was found among the four terminal alpha-amino groups of carbonmonoxyhemoglobin.

The aim of this paper was to measure the binding of CO to myoglobin and hemoglobin at various PO2 values. For this purpose we have studied an "in vitro" system made up of solutions of hemoglobin and myoglobin equilibrated in two connected tonometers with the same gas phase of various PO2 and PCO. The results indicate that a significant proportion of CO is released by hemoglobin and binds myoglobin at low PO2 values (approximately 2-3 Torr), in qualitative agreement with the predictions of a previous computer simulation of the "in vivo" system.

To investigate the mechanism of allosteric switching in human hemoglobin, we have studied the dissociation of the ligand (CO) from several intermediate ligation states by a stopped-flow kinetic technique that utilizes competitive binding of CO by microperoxidase. The hemoglobin species investigated include Hb(CO)4, the diliganded symmetrical species (alpha beta-CO)2 and (alpha-CO beta)2, and the di- and monoliganded asymmetrical species (alpha-CO beta-CO)(alpha beta), (alpha-CO beta)(alpha beta-CO), (alpha beta-CO) (alpha beta), and (alpha-CO beta)(alpha beta). They were obtained by rapid reduction with dithionite of the corresponding valence intermediates that in turn were obtained by chromatography or by hybridization. The nature and concentration of the intermediates were determined by isoelectric focusing at -25 degrees C. The study was performed at varying hemoglobin concentrations (0.1, 0.02, and 0.001 mM [heme]), pH (6.0, 7.0, 8.0), with and without inositol hexaphosphate. The results indicate that: (a) hemoglobin concentration in the 0.1-0.02 mM range does not significantly affect the kinetic rates; (b) the alpha chains dissociate CO faster than the beta chains; (c) the symmetrical diliganded intermediates show cooperativity with respect to ligand dissociation that disappears in the presence of inositol hexaphosphate; (d) the monoliganded intermediates dissociate CO faster than the diliganded intermediates; (e) the asymmetrical diliganded intermediates are functionally different from the symmetrical species.

Abstract: Blood transfusion is still a critical therapy in many diseases, traumatic events and war battlefields. However, blood cross-matching and storage may limit its applicability, especially in Third World countries. Moreover, haemoglobin, which in red blood cells is the key player in the oxygen transport from lung to tissues, when free in the plasma causes hypertension and renal failure. This investigation was aimed at the development of a novel haemoglobin-based oxygen carrier with low vasoactivity and renal filtration properties. Human haemoglobin was chemically conjugated with polyethylene glycol (PEG) under either aerobic or anaerobic conditions, following different chemical procedures. The resulting PEGylated haemoglobin products were characterized in terms of oxygen affinity, cooperativity, effects of protons and carbon dioxide concentration, and oxidation stability, and were transfused into rats to evaluate vasoactivity and renal filtration. A deoxyhaemoglobin, conjugated with seven PEG and seven propionyl groups, which we called Euro-PEG-Hb, did not produce profound hypertension, was 99% retained within 6 h, and exhibited oxygen binding properties and allosteric effects more similar to human haemoglobin A than the other tested PEGylated haemoglobin derivatives, thus appearing a very promising candidate as blood substitute. [Copyright &y& Elsevier]