WO2009000763A1

WO2009000763A1 - Process for purifying apyrogenic and virus-inactivated human hemoglobin

Info

Publication number: WO2009000763A1
Application number: PCT/EP2008/057854
Authority: WO
Inventors: Massimiliano Ramelli; Elena Della Valle; Claudio Farina; Michele Perrella; Claudia Nardini; Rodolfo Franceschini
Original assignee: Kedrion S.P.A.
Priority date: 2007-06-22
Filing date: 2008-06-20
Publication date: 2008-12-31
Also published as: ITLU20070013A1; EP2173769A1

Abstract

A process is described for purifying human hemoglobin in a form usable for the production of blood derivatives.

Description

PROCESS FOR PURIFYING APYROGENIC AND VIRUS-INACTIVATED HUMAN HEMOGLOBIN

FIELD OF THE INVENTION The invention relates to the field of blood derivatives, and in particular human hemoglobin obtained from outdated concentrated erythrocytes.

STATE OF THE ART

As is known, hemoglobin purification methods must ensure the removal of contaminants such as erythrocyte enzymes, phospholipids, surface antigens and the separation of the various forms of hemoglobin from the relevant form (HbA₀); these methods make use of the differential properties of different hemoglobin types to obtain solutions of hemoglobin which can be derivatized and/or bound to carriers, for use as acellular blood substitutes.

The development of new blood substitutes continues to arouse interest, although a finished product able to be used as an alternative to transfusion has yet to be attained.

The use of blood substitutes would enable the two main problems of transfusion medicine, namely:

- safety of the blood (absence of viruses, bacteria, prions) - lack of blood (limited number of regular donors, problems with collection) to be overcome.

A number of clinical applications have been identified where the use of blood substitutes can have a significant impact:

- replenishing blood lost during surgical operations or wounds; - replenishing iron during surgical operations or wounds;

- storing of organs destined for transplant;

- possible roles in cancer treatment.

For the production of oxygen carriers, the use of human hemoglobin solutions having undergone viral inactivation is preferred. Very often, viral inactivation is carried out by heating hemoglobin to 60-62^QC; its effectiveness is tested by "spiking" hemoglobin solutions with a model virus, such as vesicular stomatitis virus (VSV). In only a few cases has viral inactivation been verified with viruses of clinical interest such as CMV, HCV, HBV, HIV. It has been demonstrated that heat inactivation of viruses causes a 5-6 log reduction in viral load, but also leads to a substantial formation of methemoglobin.

To avoid methemoglobin formation, the heat treatment of completely deoxygenated hemoglobin or of hemoglobin in a CO atmosphere has been proposed. Unfortunately not all the data on viral inactivation have been published, despite the regulatory authorities requiring these data before going into clinical phase.

Table 1 summarizes the principal hemoglobin-based artificial carriers of oxygen and includes the main modifications, efficacy, pre-clinical and clinical profiles and the current state of development.

Clearly the first step in producing hemoglobin derivatives is to obtain a virus- inactivated hemoglobin which is purified of oxidized fractions (ferric hemoglobins) and has a low endotoxin content. Various chromatographic methods have been used for purifying hemoglobin solutions. U.S. Patent 4,925,474 by Hsia et al. describes the application of an affinity chromatography technique for purifying hemoglobin, using columns whose stationary phase possesses a ligand with high affinity for the hemoglobin binding site for cofactor DPG. For hemoglobin purification, ion exchange chromatography techniques have also been applied. The principle on which these techniques are based is well known; a mixture of different hemoglobin species in solution is applied to a suitably prepared ion exchange column. By varying the column conditions, such as solution pH, the single species can either selectively bind, or do not bind, to the column, since they have different affinities for the reactive groups in the column. Applying this technique to the purification of proteins such as hemoglobin is economically insufficient, except when used for small scale operations or for analytical work, since the quantity of hemoglobin to be bound and then eluted would be such as to require huge dimensions and very costly chromatography columns, rendering the process unachievable on an industrial level. Christensen et al., J. Biochem. Phys. 17 (1988) described a human hemoglobin purification method based on classical ion exchange chromatography which, however, cannot be adapted to productive levels for the reasons just described. Rausch and Feola (Biopure) in U.S. 5,084,558 describe anion and cation exchange chromatography methods for separating and purifying hemoglobin. In the case of anion exchange chromatography, three different approaches can be listed: a - binding the hemoglobin at high pH and eluting with a decreasing pH gradient or a decreasing pH step gradient; b - binding the hemoglobin at high pH, at low ionic strength and eluting with a saline gradient; c - loading under pH conditions such that the hemoglobin is unable to bind to the anionic resin, whereas the impurities (plus acidic contaminants) are retained by the column.

Approaches a) and b) have been amply documented but are not suitable for large scale production because the loading capacity on resin is too low to enable a sufficient resolution of the hemoglobin product. The loading capacities are generally only 20-30 mg/ml and, for a prospective industrial production, the columns required would be too costly; for example a single 50 gram dose of final product would require a 1.5 - 2.5 litre column.

Although approach c) is more feasible, it does not enable the various hemoglobin forms under consideration to be completely separated, and, above all, the percentage of methemoglobin in the purified product can reach values of up to 20%.

Pliura and Wiffen in U.S. 5,545,328 (1996), described a purification method which comprises firstly the use of an anion exchange resin to remove the more acidic contaminants from the mixture of various hemoglobin forms and subsequently the use of a cation exchange resin to remove basic contaminants. Although the authors have suppose that transferring the process to an industrial scale would be feasible, the presence of a double chromatography step leads to a series of problems to be dealt with in production, such as: the rapid regeneration of two resins with different characteristics; the availability of large volumes of two different buffer solution types; the danger of prolonging the duration of the double step through the resins, thus increasing the probability of contamination and methemoglobin formation; the operational costs being doubled relative to the costs of a process based on a single chromatography step.

In the light of the aforegiven full discussion of the current situation, the importance of being able to provide an effective method for purifying human hemoglobin is evident. DETAILED DESCRIPTION OF THE INVENTION

The present invention enables the aforesaid problems to be overcome and to obtain virus-inactivated hemoglobin in solution with a ferrihemoglobin content of less than 3%, free of stroma and cell fragments, and stable in Tris-HCI buffer for at least 6 months at temperatures less than -80^QC. This product is already a finished product available commercially for research purposes, or for companies who wish to produce haemoglobin-derived oxygen carriers.

For the production of purified hemoglobin solutions, outdated concentrated erythrocytes were used as a starting material, obtained from whole blood by centrifugation and removal of the plasma and buffy coat with subsequent resuspension of the erythrocyte concentrate in a nutrient solution. The additive solution for the erythrocytes used by us contained sodium chloride, adenine, glucose and mannitol dissolved in water, the anticoagulant solution being CPD (sodium citrate, dextrose and monobasic sodium phosphate). The initial stage of the purification process is aimed at removing the stromal component from hemoglobin. The erythrocytes are washed twice with 3 volumes of isotonic solution. The recovered pellet is hemolyzed with distilled water so as to give a hemoglobin concentration of between 7.5% and 8.5%. The hemolyzed product is centrifuged to remove all stroma and cell fragments that settle. Viral inactivation is carried out by the solvent/detergent method, preferably Triton X- 100 and TnBP.

When inactivation is complete, the inactivated hemoglobin solution is diafiltered and the diafiltrate is loaded onto a cation exchange resin equilibrated with buffer at pH of between 6.5 and 7.5, preferably 7.0, so as to enable only the ferric hemoglobin to bind to the resin, allowing the remaining forms to elute, thus separating these latter from the oxidized form (ferric hemoglobin). The hemoglobin which is not retained by the column is collected in TRIS-HCL to buffer the hemoglobin solution and render it more stable.

The process as aforedescribed enables a hemoglobin solution to be obtained which is more time stable than a more fully purified hemoglobin, as it contains, in addition to minor hemoglobins which do not influence significantly the functional properties of the product, some foreign proteins such as catalase and superoxide dismutase, which have a stabilizing effect on the product as they catalyze the destruction of oxygen radicals and hydrogen peroxide formed by the self-oxidizing reaction of hemoglobin. The process then comprises a stage of solvent/detergent viral inactivation, conducted at a sufficient temperature to limit ferric hemoglobin formation with respect to other inactivation methods, and also allows a minimal industrial production of purified hemoglobin solutions (about 20 kg to obtain 400 doses of 50 grams each), ensuring the maximum advantage in terms of process costs (column size, resin volume, buffers, consumable materials, re-use of components and reagents), logistics and production times of the product itself.

Starting from the consideration that a measure of the functional characteristics of hemoglobin is its ability to bind oxygen, the concentration of ferric hemoglobin present in solution (met-Hb%) becomes a fundamental parameter for evaluating protein stability over time. Stability studies on the finished product have shown that solutions in TRIS-HCI left at -20^QC, after one month show a relative increase in methemoglobin, this increase starting to become significant only after the first 30 days; the same solutions stored instead at -80^QC show an almost unchanged met-Hb% content (p<0.05) even 6 months after the start of the studies. As an example, results relating to the methemoglobin content of a hemoglobin batch obtained according to the method just described is given below. A sample representative of the batch was withdrawn and subdivided into aliquots, stored at -20^QC and -80^QC respectively, on which the methemoglobin content was determined at established time intervals. The relative data are given in Table 2. The thus obtained hemoglobin solution was not purified of minor hemoglobins which in any event do not influence significantly the functional properties of the product, or of foreign proteins such as catalase and superoxide dismutase, which have a stabilizing effect on the product. Therefore, an excessive purification of the red blood cell lysate was not considered expedient. For this reason, this type of process was established for a partial purification only of hemoglobin which avoids the loss of these enzyme activities. Moreover the method is convenient, if applied on an industrial scale, as the fundamental advantage of cation exchange chromatography is maintained, that is to say the advantage of achieving optimal removal of the ferric component. This component is damaging because with respect to oxygen transport it is inert; it can give rise to irreversible transformation products in the form of hemichromes; it more easily denatures and loses ferric ions which act as a catalyst in the formation of radicals, etc. This last aspect is particularly important especially when considering that hemoglobin contained in outdated erythrocytes is used as the starting material, and that during the course of the process, the hemoglobin is subjected to solvent/detergent viral inactivation treatment which, to some degree, increases the ferric component. The method of the invention is far more advantageous than classical chromatography methods proposed for hemoglobin purification in enabling the process to be implemented at the industrial level, since purification times, resin volume, column size and consequently process costs are reduced. Example 1 4 bags of outdated concentrated erythrocytes were combined; the total volume of the pooled erythrocytes was 1000 ml (tHb 16.0%, met-Hb 7.0%). The pooled erythrocytes were washed twice with 3 volumes of isotonic solution (3000 ml) at 3000 rpm for 15 minutes; at the end a pellet was recovered (970 ml). The pellet was then hemolyzed with sufficient distilled water to give a hemoglobin concentration of between 7.5% and 8.5%; in our case, therefore, 1000 ml of H₂O was added to the pellet resulting in a final volume of about 2000 ml. The hemolysate was centrifuged at 10,000 rounds for 30 minutes; stroma and cell fragments were removed in this manner, having settled at the base of the centrifuge bottle. As the volume of the recovered supernatant was 1800 ml, to prepare the solvent/detergent solution for viral inactivation, 18.4 g of Triton X-100 (1 % w/w) were weighed and dissolved in 75.7 ml of H₂O; the obtained solution was maintained under agitation at 40-50^QC. Once it had clarified, 55.4 g of TnBP (0.3% w/w) were added to the solution which was maintained under agitation for 20 minutes at ambient temperature. The solution of inactivating agents was added slowly to the supernatant which was placed under agitation at 24-26^QC, then left to agitate for 4 hours from the end of the addition.

When inactivation was complete, in order to remove the inactivating agents and to suspend the hemoglobin in the same phosphate buffer solution used for equilibrating the column, the inactivated hemoglobin solution was diafiltered with a Pellicon XL-Millipore system provided with 3 polyethersulphone (PES) membranes with a nominal filtering surface area of 50 cm², with a theoretical tangential flow rate of 30-50 ml/min and a 10 kDa cut-off (MWCO); diafiltration was conducted overnight using 20 litres (10 volumes) of phosphate buffer 7.5 mM KH₂PO₄ + 0.5 mM EDTA at pH 7.0 at 5^QC. On completion, 2000 ml of a diafiltered sample was obtained (tHb 8.0% - metHb 8.0%). Adjustment of the diafiltered product pH was unnecessary as it was found to be 6.95 (theoretical pH = 6.7).

The diafiltered sample, containing about 160 g of total hemoglobin and about 13 g of methemoglobin, was loaded onto 500 g of a cation exchange resin (CM- Cellulose), previously equilibrated with phosphate buffer 7.5 mM KH₂PO₄ + 0.5 mM EDTA at pH 7.0. That hemoglobin not retained by the column was collected in a beaker containing 300 ml of 1 M TRIS-HCI buffer at pH 8.0 to buffer and increase stability of the hemoglobin solution, whose values were: V=3500 ml; tHb 3.0%; metHb 1.5%. The solution was then concentrated to 8% with the Pellicon system (cut off: 10,000). The purification process given as an example has achieved a yield in terms of useful, i.e. non-oxidized hemoglobin, of 70.0% with a ferric hemoglobin removal capability of between 85 and 90%; furthermore, in the finished product the total hemoglobin content was 8.0%, and that of methemoglobin was 1.6%. Example 2 Example of an industrial production of 200 kg of hemoglobin for obtaining 4000 doses of 50 grams each

Since a bag of erythrocytes has an average hemoglobin content of 45 grams, 5550 bags are needed to give the required concentration. By combining 5550 bags of outdated concentrated erythrocytes, a pool with a total volume of about 1850 litres was obtained (tHb 16.0%, met-Hb 7.0%). The pool was washed twice with 5550 litres of isotonic solution at 3000 rpm for 15 minutes; at the end a pellet of about 1800 litres was recovered. The pellet (1800 litres) was hemolyzed with 1800 litres of distilled water in order to give a hemoglobin concentration comprised between 7.5% and 8.5%, giving a final volume of about 3,600 litres. The hemolysate was centrifuged at 10,000 rounds for 30 minutes. The recovered supernatant was subjected to viral inactivation with the solvent/detergent solution; as the supernatant weighed 3450 kg, 36.5 kg of Triton X-100 (1 % w/w) were weighed to be dissolved in 145 litres of H₂O. The obtained solution was placed under agitation at 40^QC - 50^QC. Once it had clarified 10.5 kg of TnBP (0.3% w/w) were added to the solution, which was placed under agitation for 20 minutes at ambient temperature. The solution of inactivating agents was added slowly to the supernatant under agitation at 24^Q - 26^QC and agitated for 4 hours from the end of the addition.

When inactivation was complete, the inactivated hemoglobin solution (V = 3650 litres) was diafiltered with a system provided with membranes having a nominal filtering surface area of 0.5 m² and with a 10 kDa cut-off; the diafiltration was carried out overnight using 18,250 litres (5 volumes) of phosphate buffer 7.5 mM KH₂PO₄ + 0.5 mM EDTA at pH 7.0 at 5^QC; on completion, 3650 litres of diafiltered sample were obtained. The pH of the diafiltrate was checked, if necessary correcting to 6.7, then it was loaded onto 400 kg of cation exchange resin (CM- Cellulose) previously equilibrated with phosphate buffer 7.5 mM KH₂PO₄ + 0.5 mM EDTA at pH 7.0. As the non-retained eluate was expected to be about 3500 litres, it was collected in 350 litres (about 1 :7 ratio) of 1 M TRIS-HCI buffer at pH 8.0 to stabilize the purified hemoglobin, whose values became: V = 3850 litres; tHb 5.5% (250 kg); metHb 1.5%. The solution may be concentrated to 8 ± 2% (10,000 cut off). Table 3 shows all the process steps, including the respective values, expressed in both percentages and kilograms, of total hemoglobin, methemoglobin, volumes and final yield of the process. The aforedescribed process enables virus inactivated hemoglobin, free of stroma and cell fragments, and with a ferrihemoglobin content of less than 3% to be obtained.

The main physico-chemical and functional parameters which characterize the thus purified hemoglobin solution are within the ranges given in Table 4.

TABLE 1

10 Hb: hemoglobin; met-Hb: methemoglobin; COP: colloidal osmotic pressure; p50: partial pressure of oxygen associated with 50% saturation of hemoglobin; T1/2: intravascular half life; ODB: outdated donor blood; r. h.: recombinant human; PEG: polyethylene glycol; n.a.: non applicable; ^* more precise information is not available (June 2004).

15

TABLE 2

TABLE 3

TABLE 4

Parameter Range p50 10-30 mmHg O₂

Hill Coefficient (n) 1.5-3.5

Cone. Haemoglobin 8.0 ± 2.0 %

Cone. Methemoglobin < 5%

Ph 7.5-8.5

Endotoxin < 0.30 EU/mg

Microbial load 0 CFU/ml

Osmolarity < 340 mOsm/kg

Pharmaceutical aspect Conforms

TnBP < 10 ppm

Triton X-100 < 10 ppm

Claims

1. Process for preparing purified hemoglobin solutions wherein outdated concentrated erythrocytes are used as the starting material, obtained from whole blood by centrifugation and removal of the plasma and buffy coat, with subsequent resuspension of the erythrocyte concentrate in a nutrient solution.

2. Process according to claim 1 wherein said nutrient solution contains sodium chloride, adenine, glucose and mannitol dissolved in water and the anticoagulant solution is CPD (sodium citrate, dextrose and monobasic sodium phosphate).

3. Process according to claims 1 and 2 wherein: - the erythrocytes are washed with an isotonic solution and lysed with distilled water;

- the hemolysate is centrifuged so as to eliminate all stroma and cell fragments that settle;

- a viral inactivation is carried out by the solvent/detergent method; - the inactivated hemoglobin solution is diafiltered and the pH of the diafiltered solution is stabilized;

- the diafiltrate is loaded onto a cation exchange resin equilibrated with phosphate buffer and the hemoglobin not retained by the column is collected;

- the hemoglobin collected in the preceding step is stabilized with TRIS-HCI buffer.

4. Process according to claim 3 wherein said erythrocyte lysis is carried out by adding distilled water in a quantity so as to bring the total hemoglobin concentration to 7.5 - 8.5%.

5. Process according to claim 3 wherein the hemolysate is centrifuged at 10,000 rounds for 30 minutes at +5^QC.

6. Process according to claim 3 wherein said inactivation is carried out with Triton X-100 and TnBP at concentrations of 1.0% and 0.3% (w/w) respectively, for a contact time of 2-16 hours at a temperature of 24^QC-26^QC.

7. Process according to claim 6 wherein the diafiltration is carried out with PES membranes with a 10 kDa cut-off and 10 volumes of phosphate buffer 7.5 mM KH₂PO₄ + 0.5 mM EDTA at pH 7.0.

8. Process according to claim 7 wherein said diafiltration is undertaken overnight at a temperature of 2^QC-8^QC.

9. Process according to claim 3 wherein the pH of the diafiltered solution is stabilized at pH 6.7.

10. Process according to claim 3 wherein said TRIS-HCI buffer has a pH of 8.0 and is in a quantity such that its concentration in the eluate is 100 mM.

1 1. Hemoglobin obtained according to the process of claims 1 -10.