J.Tze-Fei Wong Ph.D. and Jan Blumenstein, M.D., Ph.D.

A department of Biochemistry, Hong Kong University of Science & Technology, Clear Water Bay, Hong Kong, China; Dextro-Sang Corporation, Toronto, Canada

Introduction

In recent years the quest for a safe blood substitute has accelerated rapidly, propelled both by the hazardous nature of the blood supply in many parts of the world and by the increasingly satisfactory replacement of various plasma components with their recombinant counterparts. To replace the functions of erythrocytes and the bulk of plasma, a blood substitute needs to maintain effectively the oncotic pressure and volume of the circulation, and to carry oxygen from the lungs to support tissue respiration. Hemoglobin is an attractive oxygen carrier in the development of a clinical blood substitute, given its attributes as a respiratory pigment of extensive solubility, uptake and release of oxygen at appropriate partial pressures, buffering power, and above all its capability of transporting a large quantity of oxygen (Odling-Smee and Wilson 1988). However, one fundamental disadvantage of free Hb itself as a blood substitute arises from its relatively small molecular size and consequent hemoglobinuria and rapid clearance from the circulation. In view of this, the development of hemoglobin based oxygen carriers (HBOC) requires some method to slow down, or better still prevent completely, renal excretion of the Hb. A number of approaches have been utilized:

• Crosslinking the two a b-Hb dimers by chemical crosslinks in order to prevent dissociation of Hb into two halves.

• Achieve the same crosslinking by making recombinant a and/or b chains that are covalently joined together in such a way that the two a b-dimers are linked and could not dissociate.

• Polymerization of Hb to create high molecular weight polymers that could not escape into the kidney tubules.

• Joining Hb to a polymeric carrier such that the polymer-Hb conjugate is too large to escape into the kidney tubules.

• Placing the molecules into liposomal sacs resembling red blood cells, so that Hb does not come out into the plasma and enter the urine.

These approaches have succeeded to various extents in overcoming the problem of urinary excretion and short plasma half life caused by the small size of the Hb molecule. At first, it was thought that once this problem is resolved, a satisfactory HBOC could be readily obtained. Experience of the past decade, however, has made clear that the design of a satisfactory HBOC will require maximizing the performance of all molecular aspects of the HBOC. In this regard, the covalent dextran-hemoglobin (DxHb) conjugate offers important advantages.

High Yield Preparation

DxHb is prepared by conjugating human Hb to dextran, a poly1->6-α-D-glucose containing randomly distributed 1->3-α-glucopyranosyl branch residues (Yalpani 1988). Since dextran has long been used as a clinical plasma volume expander, its biocompatibility is well established. Its selection as carrier polymer in an HBOC is further suggested by its beneficial action on blood flow. The permeability from plasma to lymph of dextrans of MW8,000-500,000 daltons decreases with molecular weight (Muranishi1991), and dextrans of less than MW51,300 are not immunogenic in man (Kabat and Bezer1958). Dextran is completely metabolized or excreted from the body after brief storage in the cells of the reticuloendothelial system, and it can be chemically modified by a variety of methods to form defined and stable compounds. Indeed, the combination of water solubility, availability in a wide range of molecular sizes, and lack of significant toxicity or tissue tropism renders dextran an excellent drug carrier among biodegradable polymers. Dextran conjugation has been employed to prolong the plasma half lives of asparaginase (Wileman, Foster and Elliott1986), carboxypeptidase (Melton et al 1987), adenosine deaminase (Rosemeyer, Kornig and Seela 1982) and arginase (Sherwood et al 1977). The coupling of proteins to dextran is therefore an important means to enhance their therapeutic efficacies.

DxHb is made by coupling Hb to bromo-Dx in a high yield conjugation (Xue and Wong 1994). Since the coupling is conducted under room air without any need for cumbersome deoxygenation, it may be readily prepared on a large scale.

Protection of Kidneys

The conjugation of Hb to dextran prevents both excretion of Hb and kidney damage. When DxHb synthesized from three different sizes of Dx (MW 20,000, 40,000 and 70,000) are infused into rabbits, they do not enter into the kidneys, and their clearance rates do not vary with the size of the Dx moiety over this range (Blumentsein et al 1978). The non-excretion is not the result of interference of renal function by DxHb, because inulin excretion remains unimpaired in its presence.

In the rat model, infusion of stroma free Hb causes a marked decrease in glomerular filtration rate, and an extensive elevation in urinary N-acetyl- b-D-glucosaminidase (NAG) activity as an indicator of structural damage to renal tubular cells caused by the passage of Hb into the tubules. In contrast, DxHb, because of its larger molecular size, does not pass through the glomeruli into the tubules and leads only to a minor appearance in urinary NAG with no significant impairment in glomerular flltration rate (Tam and Wong 1988).

Non-Entry into Lymph

Importantly, conjugation to dextran also retards the non-rena1 c1earance of hemoglobin. When free Hb is infused into the body, it is cleared from the circulation with a t 1/2 of only 45 minutes, but despite the massive hemoglobinuria, less than 40% of the Hb is cleared via the renal route (Tam et al 1976). Therefore there occurs a rapid non-renal clearance of Hb. Non-renal escape of Hb into extravascular space is of utmost clinical significance for it may aggravate tissue edema during shock and other pathological states. It is noteworthy in this regard that different HBOCs, albeit renally non-excretable, vary in their circulatory residence time as a result of their different rates of non-renal clearance. It is suggested that a long plasma residence time is an essential requirement for a satisfactory HBOC in order to provide assurance that non-renal extravasation compounding tissue edema would be low. Long residence time also minimizes the need to top up the recipient frequently to maintain plasma Hb level, limits tissue damage induced by iron overload, and cuts down on the cost of HBOC treatment. Because DxHb does not enter into the lymph (Tsai and Wong 1997), its physical plasma half life is 58 hours in dogs (Tam et al 1978), one of the longest among different forms of HBOC. The fact that DxHb can function as an HBOC containing only 6% conjugated Hb further reduces both iron overload and costs.

Controllable Right-Shifting

DxHb binds and releases oxygen reversibly with an affinity that is 2.5-fold higher than that of Hb, with an in vivo half-saturating oxygen partial pressure (P 50) of 7 mm Hg (Tam, Blumenstein and Wong 1976). The oxygen affinity can be decreased by the covalent attachment of the right-shifting agent oxidized inositol tetrakisphosphate (oxyIP4). The oxyIP4 is readily obtained from partial phytase digestion of inositol hexakisphosphate, or phytate, followed by periodate oxidation to yield the dialdehyde. Covalent attachment of the dialdehyde to DxHb is achieved through reductive alkylation with dimethylamine borane as reducing agent (Wong 1988). Since the attachment effectively abolishes further right-shifting response of the Hb molecule to phytate, oxyIP4 evidently is attached to the same polyphosphate site as phytate (Xue, Wu and Wong 1992). The P 50 of oxy-IP4 right-shifted DxHb is 23 mm Hg. Because of the reduced Hill coefficient of DxHb relative to erythrocytes, during oxygen unloading from 100% down to 50% saturation, which is the normal operating range of unloading in vivo, the oxygen dissociation curve for right-shifted DxHb will be slightly right-shifted relative to the erythrocyte curve over much of this range. By controlling the percentile oxy-IP4 modification, the P 50 of DxHb may be varied continuously between 7 and 23 mm Hg to fit optimally the application of the DxHb. Thus a low oxygen affinity may be expected to expedite oxygen release by the HBOC to the tissues, and a high oxygen affinity may be expected to reduce venous oxygen tension and potential free radical toxicity of the HBOC (Alayash 2000).

Enhanced Physical Stability

Covalent coupling of Hb to dextran reduces its affinity for haptoglobin (Tam and Wong 1980). It also enhances the stability of Hb against acid denaturation and ethanol precipitation. Unike free Hb, which begins to precipitate when ethanol concentration exceeds 10%, DxHb only begins to precipitate at 30% ethanol. This opens the way to possible sterilization of DxHb solutions with organic solvents. The enhanced physical stability of DxHb provides extra safeguard against the precipitation and denaturation of the hemoglobin in the bloodstream causing circulatory obstruction. That the autooxidation of hemoglobin is slowed by 30% when complexed to dextran (Wong 1988) also enhances the utility of DxHb as HBOC.

Exchange Transfusion

When dogs were exchange transfused with either 10% dextran or 6% DxHb (with respect to Hb moiety) in kidney dialysis fluid, the animals exchange transfused with dextran could not survive when hematocrit was reduced to the 6-10% range because of lack of oxygen-delivering capacity. In contrast, the hematocrit could be lowered to 2% or less in the DxHb exchange transfused dogs, whereby the animals depended entirely on oxygen delivery by DxHb for life support. The infused DxHb exhibited a physical plasma t 1/2 of 58 hours and a functional plasma t 1/2, after correcting for non-functional Dx-metHb, of 46 hours. Because of the prolonged functional half life of DxHb in circulation, and the rapid erythropoiesis following the blood replacement, the animals went on to complete recovery under room air without any need for further transfusion. The rapid erythropoiesis observed indicates that the bone marrow was being supported vigorously by DxHb following the blood replacement. Since the bone marrow is one of the most fragile tissues in the body, with its populations of rapidly dividing cells being particularly susceptible to injury by radiation and chemicals, the health of this tissue constitutes strong evidence of the ability of DxHb to support cell proliferation and function.

Hemorrhagic Shock

While the exchange-transfusions carried out with dogs demonstrated the capacity of 6% DxHb to support life at less than 2% hematocrit, it is important also to examine its ability to do so under conditions of hemorrhagic shock. Accordingly, Guinea pigs were subjected to hemorrhagic shock and resuscitated with DxHb. The results showed that DxHb was capable of maintaining life under both exchange transfusion and hemorrhagic shock conditions (Table 1).

Table 1.

Resuscitation with DxHb of Guinea pigs in hemorrhagic shock. Guinea pigs anesthetized with pentobarbital i.p. were bled to 70% of maximum bled out over a period of 10 minutes, during which time the blood pressure dropped from 85 to 30 mm Hg. Subsequently, 0.2-0.5 ml was bled occasionally for the following 80 minutes to keep the blood pressure at 25-30 mm Hg. At 90 minutes, a bled volume of resuscitating fluid was infused over 60 minutes. Only those animals with a 90-minute lactate level between 50-90 mg/dl were included in the tests. (S.P. Tsai and J.T. Wong, unpublished results)

Conclusion

Dextran-hemoglobin is the first soluble blood substitute to sustain, without any need for further transfusion, complete recovery under room air of animals that have undergone essentially complete replacement of erythrocytes. This is made possible by its relatively long residence time in plasma, such that its disappearance from the bloodstream is adequately compensated for by the vigorous erythropoiesis it supports. The long residence time is in turn the result of non-clearance through the renal route, which protects the integrity of kidney function and structure, as well as slow clearance through non-renal routes including extravasation and cellular uptake processes. Slow extravasation would usefully limit tissue edema. The long residence time also reduces the frequency of HBOC infusions required to maintain an adequate hemoglobin level in blood. This reduction together with the relatively low concentration (6%) of Hb-moiety required by a DxHb HBOC achieve economy with regard to both the metabolic iron load imposed on the recipient and the cost of the HBOC, which is expected to be a critical factor determining the usefulness of the HBOC in many countries of the world.