ARTIFICIAL CELLS, BLOOD SUBSTITUTES, AND BIOTECHNOLOGY Vol. 32, No. 2, pp.

243262, 2004

Hemoglobin Polymerized with a Naturally Occurring Crosslinking Agent as a Blood Substitute: In Vitro and In Vivo Studies
Wen-Hsiang Chang,1 Yen Chang,2 Yi-Chien Chen,1 and Hsing-Wen Sung1,*
Department of Chemical Engineering, National Tsing Hua University, Hsinchu, Taiwan, R.O.C. 2 Division of Cardiovascular Surgery, Veterans General HospitalTaichung and College of Medicine, National Yang-Ming University, Taipei, Taiwan, R.O.C.
1

ABSTRACT
A naturally occurring crosslinking agent, genipin, extracted from the fruits of Gardenia jasminoides Ellis was used by our group to chemically modified biomolecules. Genipin and its related iridoid glucosides have been widely used as an antiphlogistic and cholagogue in herbal medicine. Our previous study showed that the cytotoxicity of genipin is significantly lower than glutaraldehyde. The study was to investigate the feasibility of using genipin to polymerize hemoglobin

*Correspondence: Hsing-Wen Sung, Department of Chemical Engineering, National Tsing Hua University, Hsinchu, Taiwan, 30013; Fax: 886-3-572-6832; E-mail: hwsung@che.nthu.edu.tw. 243
DOI: 10.1081/BIO-120037830 Copyright & 2004 by Marcel Dekker, Inc. 1073-1199 (Print); 1532-4184 (Online) www.dekker.com

ORDER

REPRINTS

244

Chang et al. as a blood substitute. The results indicated that the rate of hemoglobin polymerization by glutaraldehyde was significantly faster than that by genipin and it readily produced polymers with molecular masses greater than 500,000 Da. It was found that the maximum degree of hemoglobin polymerization by genipin was approximately 40% if over-polymerization is to be prevented. With increasing the reaction temperature, hemoglobin concentration, and genipin-to-hemoglobin molar ratio, the duration taken to achieve the maximum degree of hemoglobin polymerization by genipin became significantly shorter. The P50 value of the unmodified hemoglobin was 9 mmHg, while that of the genipin-polymerized PLP-hemoglobin increased to 21 mmHg. It was found in a rat model that the genipinpolymerized PLP-hemoglobin resulted in a longer circulation time than the unmodified hemoglobin. In conclusion, the results of the study indicated that the genipin-polymerized hemoglobin solution has a lower oxygen affinity and a longer vascular retention time than the unmodified hemoglobin solution. Key Words: Stroma-free hemoglobin; Hemoglobin polymerization; Genipin; Blood substitute; Exchange transfusion.

INTRODUCTION Hemoglobin has been used as raw materials for manufacturing blood substitutes (Chang, 1992; Everse and Hsia, 1997; Powanda and Chang, 2002). However, because of its high oxygen affinity and short vascular retention time, limitations on hemoglobin as a blood substitute in clinical therapy have been reported in the literature (Bakker et al., 1992; MacDonald and Pepper, 1994; Nelson et al., 1992). To decrease its oxygen affinity, hemoglobin has been modified by pyridoxylation and followed by polymerization with glutaraldehyde (De Venuto and Zegna, 1983; Feola et al., 1983; Lee et al., 1989; Marini et al., 1990; Sehgal et al., 1983). It was reported that the polymerized hemoglobin showed a P50 value of 1922 mmHg. Nevertheless, the reaction rate of hemoglobin with glutaraldehyde is too fast to control its molecular weight distribution (MacDonald and Pepper, 1994). Hence, polymerization of hemoglobin by glutaraldehyde is usually undertaken at 4 C. Additionally, the glutaraldehyde-polymerized hemoglobin is relatively unstable and may release glutaraldehyde residues during storage or sterilization (MacDonald and Pepper, 1994). It was reported that glutaraldehyde is cytotoxic even at low doses (MacDonald and Pepper, 1994). This may impair the biocompatibility of the polymerized products.

ORDER

REPRINTS

Genipin-Polymerized Hemoglobin

245

In an attempt to overcome the aforementioned problems, a naturally occurring crosslinking agent, genipin, was used by our group to polymerize hemoglobin. Genipin and its related iridoid glucosides extracted from the fruits of Gardenia jasminoides Ellis have been widely used as an antiphlogistic and cholagogue in herbal medicine (Akao et al., 1994). Additionally, it is known that genipin can spontaneously react with amino acids or proteins to form dark-blue pigments (Touyama et al., 1994a,b) These dark-blue pigments have been used in the fabrication of food dyes. The cytotoxicity of genipin was previously studied by our group in vitro using 3T3 fibroblasts (Sung et al., 1999). Glutaraldehyde was used as a control. The results indicated that genipin is significantly less cytotoxic than glutaraldehyde. Additionally, the genotoxicity of genipin was tested in vitro using Chinese hamster ovary (CHO-K1) cells (Tsai et al., 2000). The results hinted that glutaraldehyde may produce a weakly clastogenic response in CHO-K1 cells. In contrast, genipin does not cause clastogenic response in CHO-K1 cells. The biocompatibility of the genipin-crosslinked biological tissue was studied in a growing rat model subcutaneously (Chang et al., 2002). It was noted that the degree of inflammatory reaction for the genipin-crosslinked tissue was significantly less than its glutaraldehydecrosslinked counterpart. This implied that genipin may form biocompatible crosslinked products. The present study was to investigate the rate of hemoglobin polymerization by genipin. Glutaraldehyde was used as a control. Additionally, the effects of temperature, hemoglobin concentration, and genipin-to-hemoglobin molar ratio on the degree of hemoglobin polymerization by genipin were examined. Subsequently, the in vitro characteristics of the unmodified hemoglobin and genipin-polymerized hemoglobin solutions used for exchange transfusion were determined. Finally, the in vivo performance of the unmodified hemoglobin and genipin-polymerized hemoglobin solutions was tested in a rat model.

MATERIALS AND METHODS Preparation of Stroma-Free Hemoglobin Solution Porcine stroma-free hemoglobin solution was prepared by the aqueous two-phase system described in the literature (Hart and Bailey, 1991; Middaugh and Lawson, 1980). The porcine blood was collected from a local slaughterhouse into glass bottles containing sodium citrate solution (3.7 g/dl). The bottles were kept on ice to minimize the formation

ORDER

REPRINTS

246

Chang et al.

of methemoglobin. Upon return, the plasma was removed via centrifugation at 5,000 g for 10 min. The red blood cells were washed three times with normal saline (1:3 v/v) and lysed by treatment with three volumes DI water, a hypotonic solution, over night. Subsequently, the cell membrane remnants were removed via centrifugation at 15,000 g for 1 h. The separation and purification of stroma-free hemoglobin was performed by the aqueous two-phase system. The resulting solution was dialyzed three times against a 0.05 M phosphate buffered saline (PBS, pH 7.4), and the hemoglobin was concentrated by ultrafiltration to 10 grams per deciliter. The solution was subsequently sterilized by filtration through a 0.22-mm Millipore filter. The purity of stroma-free hemoglobin was checked by electrophoresis in SDS-polyacrylamide gels (PhastSystemTM, Pharmacia Biotech, Uppsala, Sweden) and by the gel filtration analysis using a high-performance liquid chromatographer (HPLC) equipped with a TSK G3000SWXL column (Tosoh Corp., Tokyo, Japan).

Pyridoxylation Pyridoxylation of hemoglobin was achieved by the method of Benesch et al. (1972). For a typical preparation of pyridoxal-50 -phosphatehemoglobin (PLP-Hb), 9.3 mmol (6 g/dL) of stripped hemoglobin in 10 mL of 0.1 M Tris buffer (pH 7.5 at 4 C) was deoxygenated by bubbling nitrogen through the solution, which contained 20 mL of caprylic alcohol to prevent foaming in the subsequent steps. Subsequently, 37.2 mmol of PLP was added. After 30 min under nitrogen, 186 mmol of sodium borohydride in 0.5 mL of 103 M NaOH was introduced for 30 min and then was dialyzed against isotonic PBS to remove the excess PLP and sodium borohydride. All operations were conducted at 4 C. Finally, the obtained PLP-Hb solution was concentrated by ultrafiltration to 20 g per deciliter.

Polymerization In the study, the rate of hemoglobin polymerization by genipin (Challenge Bioproducts Co., Taichung, Taiwan) was investigated. Glutaraldehyde was used as a control. Additionally, the effects of temperature (4, 10, 25 C), hemoglobin concentration (PLP-Hb concentration in 2, 6, 10 g/dl), and genipin-to-hemoglobin molar ratio (3/1, 5/1, 7/1, 10/1) on the degree of hemoglobin polymerization by genipin

ORDER

REPRINTS

Genipin-Polymerized Hemoglobin

247

(GP-PLP-Hb) were investigated. The degree of hemoglobin polymerization by genipin was monitored by the gel filtration analysis with a TSK G3000SWXL column. In order to prevent the formation of polymeric Hb (GP-PLP-Hb) with a molecular weight greater than 500,000 Da, the polymerization reaction was terminated by the addition of glycine. It is known that genipin can spontaneously react with glycine (Fujikawa et al., 1987). Consequently, the effect of using glycine at a variety of concentrations (in glycine-to-hemoglobin molar ratio) on the termination of hemoglobin polymerization by genipin was examined. Subsequently, the polymerized GP-PLP-Hb solution was dialyzed to eliminate any unreacted genipin, glycine, and polymerized glycine. Finally, the obtained GP-PLP-Hb solution was concentrated by ultrafiltration to 10 grams per deciliter. Total hemoglobin and methemoglobin concentrations were measured as per the methods described by Crosby and co-workers and Evelyn and Malloy, respectively (Crosby et al., 1954; Evelyn and Malloy, 1938).

Removal of Unpolymerized Hemoglobin A final consideration for clinical products is the need to reduce the residual unpolymerized hemoglobin to a minimum (MacDonald and Pepper, 1994). In the study, the removal of unpolymerized hemoglobin was carried out by an ion-exchange column (DEAE cellulose column, Sigma Chemical Co., St. Louis, Missouri, USA) or a gel-filtration column (Sephadex, G-100-120, Sigma Chemical Co.).

Characteristics of Test Solutions for Exchange Transfusion The characteristics of the unmodified Hb and GP-PLP-Hb solutions used for exchange transfusion in the rat were determined as follows: the sodium and potassium concentrations by an atomic absorption spectrophotometer (Model AA-100, Perkin Elmer Inc., Norwalk, Connecticut, USA) and the osmolality values by an osmometer (Advanced Micro Osmometer, Model 3300, Norwood, Massachusetts, USA). Hemoglobin oxygen affinity measurements were obtained using the biotonometry technique reported by Neville (1974). The distributions in particle size for the unmodified Hb and GP-PLP-Hb solutions (n 3) were analyzed by a light-scattering method (ZetaSizer, Trekintal Corp., Taipei, Taiwan).

ORDER

REPRINTS

248

Chang et al.

Animal Studies Animal studies were conduced in two groups of male Sprague-Dawley rats. The first group of experiments was to examine the survival of the rats at an approximately 50% blood-volume-exchange transfusion with the GP-PLP-Hb solution. The unmodified Hb and PBS were used as controls. Six rats (200250 g) were used for each exchange transfusion study. The animals were anesthetized and the femoral vein was cannulated with a polyethylene catheter connected by a three-way sterile stopcock to two syringes (Lee et al., 1989), one used for phlebotomy and the other for infusion of sample solution (De Venuto et al., 1977). Exchange transfusion was done by removal one ml of blood and immediate infusion of one ml of sample solution, repeating the process until the desired exchange level was achieved. The percentage of blood-volume-exchange was calculated assuming that the rat blood volume was 10% of its body weight. The second group of experiments was to test the half-life of the GP-PLP-Hb in blood circulation. The unmodified Hb was used as a control. The experimental preparation and procedure were the same as aforementioned. Samples of blood were obtained from the tail vein at various intervals during the posttransfusion hours.

Statistical Analysis Statistical analysis for the determination of differences in the measured properties between groups was accomplished using one-way analysis of variance and determination of confidence intervals, performed with a computer statistical program (Statistical Analysis System, Version 6.08, SAS Institute Inc., Cary, North Carolina, USA). All data are presented as a mean value with its standard deviation indicated (mean SD).

RESULTS Stroma-Free Hemoglobin Solution The purified stroma-free hemoglobin solution was obtained by the aqueous two-phase system. A typical elution pattern from the TSK G3000SWXL column for the purified stroma-free hemoglobin solution is shown in Fig. 1a. The stroma-free hemoglobin solution was eluted from the column as a single peak corresponding to a molecular weight of

ORDER

REPRINTS

Genipin-Polymerized Hemoglobin

249

Figure 1. (a) A typical elution pattern from the TSK G3000SWXL column for the purified stroma-free hemoglobin solution; (b) Electrophoresis in SDS-page of the purified stroma-free hemoglobin solution.

64,000 Da. Electrophoresis in SDS-page of the purified stroma-free hemoglobin solution yielded a single band corresponding to the 16,000dalton monomer (Fig. 1b). These results demonstrated that the stromafree hemoglobin solution obtained by the aqueous two-phase system was highly pure.

Polymerized Hemoglobin Solution A typical HPLC molecular weight profile following 4 h after initiation of polymerization of hemoglobin (10 g/dL of PLP-Hb) by genipin (7/1 molar ratio of GP/Hb ) at 15 C is shown in Fig. 2a. The unpolymerized fraction (PLP-Hb) was approximately 60% and there was very little polymeric GP-PLP-Hb with a molecular weight greater than 500,000 Da. In comparison, a typical HPLC molecular weight profile following 30 min after initiation of polymerization of hemoglobin by glutaraldehyde at the same reaction conditions is show in Fig. 2b. As indicated in the figure, the rate of hemoglobin polymerization by glutaraldehyde was significantly faster than that by genipin (P < 0.05) and it readily produced polymers with molecular masses greater than 500,000 Da. It was found that the maximum degree of hemoglobin polymerization by genipin was approximately 40% if the formation of polymeric

ORDER

REPRINTS

250
45 40 35 30 mv 25 20 15 10 5 25 20 15 10 5 45 40 35 30 mv

Chang et al.

10 15 minutes

20

25

(a)
20 20

15

over polymerization

15

mv 10

mv 10

10 15 minutes

20

25

(b)
Figure 2. (a) A typical HPLC molecular weight profile following 4 h after initiation of polymerization of hemoglobin by genipin; (b) A typical HPLC molecular weight profile following 30 min after initiation of polymerization of hemoglobin by glutaraldehyde at the same reaction conditions.

ORDER

REPRINTS

Genipin-Polymerized Hemoglobin

251

GP-PLP-Hb with a molecular weight >500,000 Da is to be prevented. This maximum degree of hemoglobin polymerization by genipin could be achieved at different reaction conditions at distinct durations. The durations taken for various reaction conditions (temperature, hemoglobin concentration, and genipin-to-hemoglobin molar ratio) to achieve the maximum degree of hemoglobin polymerization by genipin (40%) were listed in Table 1. As shown in the table, all reaction conditions investigated significantly influenced the rate of hemoglobin polymerization by genipin. With increasing the reaction temperature, hemoglobin concentration, and genipin-to-hemoglobin molar ratio, the duration taken to achieve the maximum degree of hemoglobin polymerization by genipin became significantly shorter (P < 0.05). Table 2 presents the percentage of methemoglobin produced at each corresponding reaction condition investigated after the maximum degree of hemoglobin polymerization by genipin was achieved. The results indicated that the percentages of methemoglobin observed at all investigated conditions were minimal, even at room temperature (25 C). The conditions used to polymerize hemoglobin by genipin for the following studies were: a hemoglobin (PLP-Hb) concentration of 10 g/dL, a genipin-to-hemoglobin molar ratio of 7/1, and a reaction temperature at 15 C. The corresponding HPLC molecular weight profile for the genipin-polymerized hemoglobin (GP-PLP-Hb) solution under such conditions is already shown in Fig. 2a.
Table 1. The durations taken (h) for various reaction conditions (temperature, hemoglobin concentration, and genipin-to-hemoglobin molar ratio) to achieve the maximum degree of hemoglobin polymerization by genipin (40%). Genipin/Hb (molar ratio) 25 C Hb conc. 10 g/dL 6 g/dL 2 g/dL Hb conc. 10 g/dL 6 g/dL 2 g/dL Hb conc. 10 g/dL 6 g/dL 2 g/dL

3/1 3 3.5 8.5 6 12 24 12 14 32

5/1 2 2.5 5.5 4 8 16 6 10 24

7/1 2 2 4.5 4 4 12 4 8 18

10/1 1 2 3 2 4 12 4 6 14

15 C

10 C

ORDER

REPRINTS

252

Chang et al.

Table 2. Percentages of methemoglobin produced at various reaction conditions investigated (temperature, hemoglobin concentration, and genipin-to-hemoglobin molar ratio) after the maximum degree of hemoglobin polymerization by genipin was achieved. Genipin/Hb (molar ratio) 25 C Hb conc. 10 g/dL 6 g/dL 2 g/dL Hb conc. 10 g/dL 6 g/dL 2 g/dL Hb conc. 10 g/dL 6 g/dL 2 g/dL

3/1 1.0 0.2 2.6 0.3 1.2 0.2 3.2 0.3 1.1 0.2 2.1 0.2 1.4 0.1 1.4 0.2 1.5 0.1

5/1 2.6 0.1 1.5 0.1 1.0 0.1 1.2 0.1 1.2 0.1 1.7 0.1 2.7 0.3 1.4 0.1 1.0 0.1

7/1 2.3 0.3 1.2 0.2 0.8 0.1 1.3 0.1 1.6 0.3 1.6 0.2 1.8 0.2 1.6 0.2 1.4 0.2

10/1 3.2 0.1 1.2 0.3 1.2 0.1 1.1 0.2 2.4 0.2 1.1 0.2 2.5 0.4 3.3 0.4 1.6 0.2

15 C

10 C

In order to prevent over polymerization (the formation of polymeric GP-PLP-Hb with a molecular weight >500,000 Da), the hemoglobin polymerization by genipin has to be terminated. In the study, the hemoglobin polymerization by genipin was terminated by the addition of glycine. The effect of using glycine at various concentrations (in glycineto-hemoglobin molar ratio) on the termination of hemoglobin polymerization by genipin was shown in Fig. 3. The molecular weight profile of the genipin-polymerized hemoglobin (GP-PLP-Hb) shifted slightly towards higher molecular weight (compared with the result presented in Fig. 2a), an indication of over polymerization, when the glycine-tohemoglobin molar ratio used was 20/1 or 50/1. This phenomenon in over polymerization could be effectively prevented if the glycine-to-hemoglobin molar ratio was greater than 100/1. Under this condition, there was no further buildup of higher molecular weight polymers.

Removal of Unpolymerized Hemoglobin The results of removal of the unpolymerized hemoglobin (PLP-Hb) from the GP-PLP-Hb carried out by an ion-exchange column or a

ORDER

REPRINTS

Genipin-Polymerized Hemoglobin

253

Glycine/Hb
70 60 mv 50 40 30 20 10 5

20/1
80 70 60 50 40 30 20 10 10 15 minutes 20 25 mv mv 70 60 50 40 30 20 10 5

50/1
80 over polymerization 80 70 60 50 40 30 20 10 10 15 minutes 20 25 mv

80 over polymerization

100/1
80 70 60 mv 50 40 30 20 10 5 10 15 minutes 20 80 70 60 40 50 mv mv 30 40 30 20 10 25 20 10 5 60 50

200/1
60 50 40 mv 30 20 10 10 15 minutes 20 25

Figure 3. Effects of using glycine at various concentrations (in glycine-tohemoglobin molar ratio) on the termination of hemoglobin polymerization by genipin.

gel-filtration column were presented in Fig. 4. As shown in the figure, after passage through the ion-exchange column, the fraction of residual unpolymerized hemoglobin was reduced from 60% (Fig. 4a) to 27% (Fig. 4b). In contrast, there was very little unpolymerized hemoglobin left in its HPLC molecular weight profile after passage through the gel-filtration column (Fig. 4c). This indicated that removal of the unpolymerized hemoglobin from the genipin-polymerized hemoglobin by the gel-filtration column was significantly more effective than by the ionexchange column (P < 0.05).

ORDER

REPRINTS

254

Chang et al.

45 40 35 30 mv 25 20 15 10 5 5 10 15 minutes 20

45

Unpolymerized 40 PLP-Hb (60%)

35 30 25 20 15 10 5 25 mv

65

65

60

Unpolymerized PLP-Hb (27%)

60

55

55

50

50

45

10

15 minutes

20

45 25

(a)
10 10

(b)

mv 7.5

7.5

5 5 10 15 minutes 20

5 25

(c)

Figure 4. Results of removal of the unpolymerized hemoglobin (PLP-Hb) from the genipin-polymerized hemoglobin (GP-PLP-Hb) carried out by an ion-exchange column (b) or a gel-filtration column (c).

Characteristics of Test Solutions for Exchange Transfusion The characteristics of the unmodified Hb and GP-PLP-Hb solutions used for exchange transfusion in the rat are shown in Table 3. The two solutions were essentially the same in their compositions except for the parameter of P50 value. The P50 values of the Hb and GP-PLP-Hb solutions were 9 mmHg and 21 mmHg, respectively. The distribution curves in particle size for the unmodified Hb and GP-PLP-Hb solutions determined by a light-scattering method are presented in Fig. 5. As shown in the figure, the distribution in particle size for the GP-PLP-Hb solution was significantly wider than that for the unmodified Hb solution. Additionally, the averaged particle size for the GP-PLPHb solution (32.0 7.4 nm) was significantly larger than the unmodified Hb solution (6.8 0.3 nm, P < 0.05).

ORDER

REPRINTS

Genipin-Polymerized Hemoglobin

255

Table 3. Characteristics of the unmodified Hb and GP-PLP-Hb solutions used for exchange transfusions in the rat. Hb Concentration (g/dL) MetHb (%) Na (meq) K (meq) Osmolality (mOsm) PH P50 (mmHg) 10 3.3 151 4.6 287 7.4 9 GP-PLP-Hb 10 3.0 136 4.4 289 7.4 21

Unmodified Hb

GP-PLP-Hb

6.8 0.3 nm

32 7.4 nm

Figure 5. Particle-size distribution curves for the unmodified Hb and GP-PLPHb solutions determined by a light-scattering method. (View this art in color at www.dekker.com.)

Animal Studies Test solutions for exchange transfusion in the rat were sterilized by filtration through a 0.22-mm Millipore filter. The results of the first group of experiments in examining the survival of the rats at an approximately 50% blood-volume-exchange transfusion with the unmodified Hb, PBS, or GP-PLP-Hb solution are summarized in Table 4. It was found that

ORDER

REPRINTS

256

Chang et al.

Table 4. Results of the first group of experiments in examining the survival of the rats at an approximately 50% blood-volume-exchange transfusion with the unmodified Hb, PBS, or GP-PLP-Hb solution. Percentage of exchange Hb PBS GP-PLP-Hb 53.3 0.9 54.3 6.5 51.7 1.3 Survival ratio (n 6) 1/6 1/6 6/6 Averaged survival period (h) 5.7 1.3 3.6 0.4 >3 months

4.5

Plasma Hb Concentration (g/dl)

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 0.0 2.5 5.0 7.5

(n = 6)

Hb GP-PLP-Hb

10.0 12.5 15.0 17.5 20.0 22.5 25.0

Time (hours)

Figure 6. Disappearance of plasma hemoglobin for the animals treated with the unmodified Hb or GP-PLP-Hb solution.

the results for the animals treated with the GP-PLP-Hb solution were significantly superior to the other control groups (P < 0.05). In this specific test group, all test animals (n 6) after treatment with the GPPLP-Hb solution survived and remained healthy more than 3 months. In contrast, only one of six rats survived for the control groups treated with the unmodified Hb or PBS solution, while the other test animals in the corresponding groups died in about 5 h after exchange transfusion. The results of the second group of experiments in testing the half-life durations of the unmodified Hb and GP-PLP-Hb solutions in circulation are given in Fig. 6. As shown in the figure, the disappearance of hemoglobin was significantly faster in the animals treated with the unmodified Hb solution than their counterparts treated with the GPPLP-Hb solution (P < 0.05). The half-life duration, that is the time

ORDER

REPRINTS

Genipin-Polymerized Hemoglobin

257

necessary for the plasma hemoglobin concentration to reach one half the value observed at the end of exchange transfusion, was 1.5 h for the control group treated with the unmodified Hb solution and 12.5 h for the test group treated with the GP-PLP-Hb solution.

DISCUSSION It is well known that high oxygen affinity and short vascular retention time represent two important limitations for the stroma-free hemoglobin solution in the blood-volume-exchange transfusion (De Venuto and Zegna, 1983; Keipert and Chang, 1985). To overcome the former limitation, an agent to intramolecularly crosslink stroma-free hemoglobin has long been thought desirable. Intramolecular crosslinks between tetramer subunits prevent dissociation into excretable dimmers (32,000 Da). Benesch et al. demonstrated that PLP can be attached to the N-terminal valine of hemoglobin  chains by forming a Schiff base and that the resulting PLP-Hb has lower oxygen affinity than the unmodified Hb (Benesch et al., 1972). It was reported that pyridoxylation of hemoglobin appeared to be necessary to maintain the cooperativity of the molecular chains and the ability to reversibly bind oxygen when hemoglobin is subsequently subjected to polymerization (De Venuto and Zegna, 1983). To overcome the latter limitation, it was reported that intermolecular polymerization of PLP-Hb can yield a product with a longer vascular retention time and at the same time with a lower oxygen affinity than the unmodified Hb (Bakker et al., 1992; Keipert and Chang, 1985). However, carrying out intermolecular polymerization too far, producing polymers with molecular masses > 500,000 Da, is undesirable because the large aggregates may have altered surface charge characteristics (MacDonald and Pepper, 1994). These alterations could lead to flow changes in the microcirculation and may well acquire additional untoward toxicities, such as reticuloendothelial system blockade or stimulation of an immunological response to the altered aggregate surface (MacDonald and Pepper, 1994). Glutaraldehyde is a well-known crosslinking agent, producing rapid intermolecular, as well as limited intramolecular, polymerization of hemoglobin (MacDonald and Pepper, 1994). Due to its rapid reaction, a mixture of glutaraldehyde-polymerized hemoglobin with a wide range of molecular weight is readily produced. Additionally, a reversible reaction releasing free glutaraldehyde after hemoglobin polymerization is possible (MacDonald and Pepper, 1994). It was reported that glutaraldehyde is cytotoxic even at low doses (MacDonald and Pepper, 1994). The

ORDER

REPRINTS

258

Chang et al.

mechanism of polymerization of hemoglobin by glutaraldehyde can be found in the literature (MacDonald and Pepper, 1994). It was found in our previous study that genipin can react with the free amino groups of lysine, hydroxylysine, or arginine residues within biological tissues (Sung et al., 1998). Touyama et al. studied the structures of the intermediates leading to a blue pigment produced from genipin and methylamine, the simplest primary amine (Touyama et al., 1994a, 1994b). The presumed mechanism for the formation of the genipin-methylamine monomer and the blue-pigment polymers proposed by the Touyamas group can be found in the literature (Touyama et al., 1994a,b). Briefly speaking, the genipin-methylamine monomer is formed through a nucleophilic attack by methylamine on the olefinic carbon at C-3 of genipin, followed by opening of the dihydropyran ring and attacked by the secondary amino group on the resulting aldehyde group (Touyama et al., 1994a, 1994b). The blue-pigment polymers are presumably formed through oxygen radical-induced polymerization and dehydrogenation of several intermediary pigments. The results of the aforementioned studies suggest that genipin may form intermolecular polymerization of hemoglobin (Fig. 7) (Fujikawa et al., 1987, 1988). It was found that the stability of the genipin-crosslinked tissue during storage was superior to its glutaraldehyde-crosslinked counterpart (Sung et al., 2001). The differences in stability between the genipin- and glutaraldehyde-crosslinked tissues during storage may be caused by their

O C 11 5 7 9 CH2 10

OCH3
COOCH3

COOCH3
CH3

3 O 1 OH
OH H

: NH2

.. NH O

OH

OH

Genipin (GP)

hemoglobin

O C CH3

OCH3

GP

N+ H3CO C O CH H

GP

GP

CH3

Figure 7. Presumable mechanism of hemoglobin polymerization by genipin. (View this art in color at www.dekker.com.)

ORDER

REPRINTS

Genipin-Polymerized Hemoglobin

259

differences in crosslinking structure. It is known that glutaraldehyde may react with free amino groups to form Schiff-bases. It was reported that the reaction of Schiff-base is relatively unstable and reversible (Carey, 1992; Sung et al., 2001). On the other hand, genipin may react with free amino groups and form a tertiary amine structure which is more stable than Schiff-base (Fig. 7) (Carey, 1992). The results shown in our study revealed that the rate of hemoglobin polymerization by glutaraldehyde was significantly faster than that by genipin (Figs. 2a and b). It was reported in the literature that polymerization by glutaraldehyde is a rapid reaction (MacDonald and Pepper, 1994). Therefore, it readily produced glutaraldehyde-polymerized hemoglobin with molecular masses greater than 500,000 Da (Fig. 2b). In contrast, there was very little polymeric GP-PLP-Hb with a molecular weight >500,000 Da observed (Fig. 2a). The rate of hemoglobin polymerization by genipin may be influenced by the conditions at which the reaction is run. It was noted in the study that with increasing reaction temperature, hemoglobin concentration, and genipin-to-hemoglobin molar ratio, the duration taken to achieve the maximum degree of hemoglobin polymerization by genipin became significantly shorter (Table 1). To prevent over polymerization, the hemoglobin polymerization by genipin can be terminated by the addition of glycine. It was found that once the addition of glycine-to-hemoglobin molar ratio was greater than 100/1, no further buildup of higher molecular weight polymers was observed (Fig. 3). The removal of the unpolymerized hemoglobin (PLP-Hb) from the genipin-polymerized hemoglobin (GPPLP-Hb) can be effectively achieved by a gel-filtration column (Fig. 4). It was found in the animal study that the GP-PLP-Hb solution resulted in a longer circulation time (i.e., a greater half life in the disappearance of plasma hemoglobin) than the unmodified Hb solution (Fig. 6). It was reported that a main route of Hb clearance is through the kidney (Bleeker et al., 1989; Lenz et al., 1991; Ning et al., 1992; Savitsky et al., 1978). Renal glomeruli filter proteins under 65,000 Da out of the circulation. Free unmodified Hb, with its tendency to dissociate into dimmers (32,000 Da), is readily cleared from the blood and excreted into the urine (Bunn et al., 1969). The averaged particle size of the GP-PLPHb measured was significantly greater than the unmodified Hb (Fig. 5). The large size of the GP-PLP-Hb may reduce its ability to pass into extra-vascular spaces. In conclusion, the results of the study indicated that the genipinpolymerized hemoglobin (GP-PLP-Hb) solution has a lower oxygen affinity and a longer vascular retention time than the unmodified Hb solution.

ORDER

REPRINTS

260

Chang et al.

ACKNOWLEDGMENTS This work was supported partly by a grant from the National Science Council of Taiwan, Republic of China (NSC91-2320-B-007-004) and partly by another grant from the Veterans General Hospital, Tsing-Hua, Yang-Ming Joint Research Program (VTY-92-P4-21).

REFERENCES Akao, T., Kobashi, K., Aburada, M. (1994). Enzymatic studies on the animal and intestinal bacterial metabolism of geniposide. Biol. Pharm. Bull. 17:15731576. Bakker, J. C., Berbers, G. A., Bleeker, W. K., Den Boer, P. J., Biessels, P. T. (1992). Preparation and characterization of crosslinked and polymerized hemoglobin solution. Biomat. Art Cells, Immob. Biotech. 20:233241. Benesch, R. E., Benesch, R., Renthal, R. D., Maeda, N. (1972). Affinity labeling of the polyphosphate binding site of hemoglobin. Biochemistry 11:35763582. Bleeker, W., Van der Plas, J., Feitsma, R., Agterberg, J., Rigter, G., DeVries-van Rossen, A., Pauwels, E., Bakker, J. (1989). In vivo distribution and elimination of hemoglobin modified by intramolecular crosslinking with 2-nor-2-formylpyridoxal50 -phosphate. J. Lab. Clin. Med. 113:151161. Bunn, H. F., Esham, W. T., Bull, R. W. (1969). The renal handling of hemoglobin. I. Glomerular filtration. J. Exp. Med. 129:909923. Carey, F. A. (1992). Organic Chemistry. 2nd ed. New York: McGraw-Hill. Chang, T. M. (1992). Blood substitutes based on modified hemoglobin prepared by encapsulation or crosslinking: an overview. Biomat. Art. Cells, Immob. Biotech. 20(24):159179. Chang, Y., Tsai, C. C., Liang, H. C., Sung, H. W. (2002). In vivo evaluation of cellular and acellular bovine pericardia fixed with a naturally occurring crosslinking agent (genipin). Biomaterials 23:24472457. Crosby, W. H., Munn, J. I., Furth, F. W. (1954). Standardizing a method for clinical hemoglobinometry. U.S. Armed. Forces Med. J. 5:693. De Venuto, F., Moores, W. Y., Zegna, A. I., Zuck, T. F. (1977). Total and partial blood exchange in the rat with hemoglobin prepared by crystallization. Transfusion 17:555562.

ORDER

REPRINTS

Genipin-Polymerized Hemoglobin

261

De Venuto, F., Zegna, A. (1983). Preparation and evaluation of pyridoxylated-polymerized human hemoglobin. J. Surg. Res. 34:205212. Evelyn, K. A., Malloy, H. T. (1938). Microdetermination of oxyhemoglobin, methemoglobin and sulfhemoglobin in a single sample of blood. J. Biol. Chem. 126:655. Everse, J., Hsia, N. (1997). The toxicity of native and modified hemoglobins. Free Radical Bio. Med. 22:10751099. Feola, M., Gonzalez, H., Canizaro, P. C., Bingham, D., Periman, P. (1983). Development of a bovine stroma-free hemoglobin solution as a blood substitute. Surg. Gynecol. Obstet. 157:399408. Fujikawa, S., Fukui, Y., Koga, K. (1987). Structure of genipocyanin G1, a spontaneous reaction product between genipin and glycine. Tetrahedron Letters 28:46994700. Fujikawa, S., Nakamura, S., Koga, K. (1988). Genipin, a new type of protein crosslinking reagent from gardenia fruits. Agric. Biol. Chem. 52:869870. Hart, R. A., Bailey, J. E. (1991). Purification and aqueous two-phase partitioning properties of recombinant vitreoscilla hemoglobin. Enzyme Microb. Tech. 13:788795. Keipert, P. E., Chang, T. M. (1985). Pyridoxylated polyhemoglobin as a red cell substitute for resuscitation of lethal hemorrhagic shock in conscious rats. Biomat. Med. Dev. Art. Org. 13:115. Lee, R., Atsumi, N., Jacobs, E. E., Austen, W. G., Vlahakes, G. J. (1989). Ultrapure, stroma-free, polymerized bovine hemoglobin solution: evaluation of renal toxicity. J. Surg. Res. 47:407411. Lenz, G., Junger, H., Scheider, M., Kothe, N., Lissner, R., Prince, A. (1991). Elimination of pyridoxylated polyhemoglobin after partial exchange transfusion in chimpanzees. Biomat. Art. Cells, Immob. Biotech. 19:699708. MacDonald, S. L., Pepper, D. S. (1994). Hemoglobin polymerization. In: Everse, J., Vandegriff, K. D., Winslow, R. M., eds. Methods in Enzymology. San Diego, CA: Academic, 231, 287308. Marini, M. A., Moore, G. L., Christensen, S. M., Fishman, R. M., Jessee, R. G., Medina, F., Snell, S. M., Zegna, A. I. (1990). Reexamination of the polymerization of pyridoxylated hemoglobin with glutaraldehyde. Biopolymers 29:871882. Middaugh, C. R., Lawson, E. Q. (1980). Analysis of protein association by partitioning in aqueous two-phase polymer system: application to the tetramerdimer dissociation of hemoglobin. Analytical Biochemistry 105:364368.

ORDER

REPRINTS

262

Chang et al.

Nelson, D., Srnak, A., Ebeling, A., Kunas, G., Catarello, J., Burhop, K. (1992). Synthesis and properties of polymerized diasprin crosslinked hemoglobins. Biomat. Art. Cells, Immob. Biotech. 20:253258. Neville, J. R. (1974). Hemoglobin oxygen affinity measurement using biotonometry. J. Appl. Physiol. 37:967971. Ning, J., Peterson, L., Anderson, P., Biro, G. (1992). Systemic hemodynamic and renal effects of unmodified human SFHS in dogs. Biomat. Art. Cells, Immob. Biotech. 20:723727. Powanda, D. D., Chang, T. M. (2002). Cross-linked polyhemoglobinsuperoxide dismutase-catalase supplies oxygen without causing blood-brain barrier disruption or brain edema in a rat model of transient global brain ischemia-reperfusion. Art. Cells, Immob. Biotech. 30(1):2337. Savitsky, J., Doczi, J., Black, J., Arnold, J. (1978). A clinical safety trial of stroma-free hemoglobin. Clin. Pharmacol. Ther. 23:7380. Sehgal, L. R., Rosen, A. L., Gould, S. A., Sehgal, H. L., Moss, G. S. (1983). Preparation and in vitro characteristics of polymerized pyridoxylated hemoglobin. Transfusion 23:158162. Sung, H. W., Huang, R. N., Huang, L. L. H., Tsai, C. C. (1999). In vitro evaluation of cytotoxicity of a naturally occurring crosslinking reagent for biological tissue fixation. J. Biomat. Sci. Polym. Edn. 10:6378. Sung, H. W., Huang, R. N., Huang, L. L. H., Tsai, C. C., Chiu, C. T. (1998). Feasibility study of a natural crosslinking reagent for biological tissue fixation. J. Biomed. Mater. Res. 42:560567. Sung, H. W., Liang, I. L., Chen, C. N., Huang, R. N., Liang, H. F. (2001). Stability of a biological tissue fixed with a naturally occurring crosslinking agent (genipin). J. Biomed. Mater. Res. 55:538546. Touyama, R., Takeda, Y., Inoue, K., Kawamura, I., Yatsuzuka, M., Ikumoto, T., Shingu, T., Yokoi, T., Inouye, H. (1994a). Studies on the blue pigments produced from genipin and methylamine. I. Structures of the brownish-red pigments, intermediates leading to the blue pigments. Chem. Pharm. Bull. 42:668673. Touyama, R., Inoue, K., Takeda, Y., Yatsuzuka, M., Ikumoto, T., Moritome, N., Shingu, T., Yokoi, T., Inouye, H. (1994b). Studies on the blue pigments produced from genipin and methylamine. II. On the formation mechanisms of brownish-red intermediates leading to the blue pigment formation. Chem. Pharm. Bull. 42:15711578. Tsai, C. C., Huang, R. N., Sung, H. W., Liang, H. C. (2000). In vitro evaluation of the genotoxicity of a naturally occurring crosslinking agent (genipin) for biological tissue fixation. J. Biomed. Mater. Res. 52:5865.