Properties of and Oxygen Binding by Albumin ... - ACS Publications

Akito Nakagawa, Teruyuki Komatsu, and Eishun Tsuchida ... Kanazawa , Sumio Tanase , Keiichi Kawai , Toru Maruyama , Ayaka Suenaga , Masaki Otagiri...
0 downloads 0 Views 158KB Size
Bioconjugate Chem. 1997, 8, 534−538

534

Properties of and Oxygen Binding by Albumin-Tetraphenylporphyrinatoiron(II) Derivative Complexes Eishun Tsuchida,*,† Katsutoshi Ando, Hiromitsu Maejima, Noriyuki Kawai, Teruyuki Komatsu, Shinji Takeoka, and Hiroyuki Nishide Department of Polymer Chemistry, Advanced Research Institute for Science and Engineering, Waseda University, Tokyo 169, Japan. Received December 23, 1996X

A hydrophobic tetraphenylporphyrinatoiron(II) derivative bearing a covalently bound axial imidazole [Fe(II)P] was efficiently and noncovalently bound into human serum albumin (HSA) up to an average of eight Fe(II)P molecules per HSA molecule. The aqueous solutions of the HSA-Fe(II)P complex provided a reversible and relatively stable oxygen adduct under physiological conditions (pH 7.4 and 37 °C). The half-life of the oxygen adduct (τ1/2) was 1 h at 37 °C in an air atmosphere. With Fe(II)TpivPP (the so-called “picket-fence heme”) having no axial base, an oxygenated HSA-Fe(II)TpivPP complex was obtained using a 20-fold molar excess of 1,2-dimethylimidazole, but the τ1/2 was very short (ca. 10 min at 37 °C). The oxygen affinity [P1/2(O2)] and oxygen transporting efficiency (OTE) of HSA-Fe(II)P at 37 °C were 30 Torr and 22%, respectively. Furthermore, the oxygen-binding and dissociation rate constants (kon and koff) are extremely high in comparison with those of hemoglobin. The HSA molecule binding eight Fe(II)P molecules can transport about 3.4 mL/dL of oxygen under physiological conditions, corresponding to about 60 % of the oxygen transporting amount of human blood.

INTRODUCTION

Human serum albumin (HSA)1 is the second major protein, following hemoglobin, of the proteins in blood. Its molecular mass is 66.5 kDa, which is similar to that of hemoglobin (64.5 kDa). HSA maintains the colloidal osmotic pressure of blood and transports nutritional and metabolic products, physiologically active compounds, toxic materials, heavy metals, and so on. For example, hemin released from hemoglobin by hemolysis is trapped by HSA in plasma and is transported to the liver for metabolic processing. The formation of the HSA-hemin complex has been widely studied (Beaven et al., 1974; Adams & Berman, 1980; Morgan et al., 1980; Lamola et al., 1981; Hrkal & Klementova, 1983; Kodicek et al., 1983; Moehring et al., 1983); however, there has been no report on the reduction of hemin to heme and the oxygenation of the heme bound to HSA. Marden et al. (1989) determined the carbonylation of a deoxyheme-HSA complex in an aqueous solution, and Bonaventura et al. (1994) preliminarily reported the reversible spectral change of HSA binding a heme derivative in presence of a large molar excess of axial imidazole upon exposure to dioxygen but did not succeed in obtaining a stable oxygen adduct. * Author to whom all correspondence should be addressed. Fax: +81 3-3209-5522. Phone: +81 3-3209-8895. †CREST investigator, JST. X Abstract published in Advance ACS Abstracts, July 1, 1997. 1 Abbreviations: Fe(II)P, 2-[[[8-[N-(2-methylimidazolyl)]octanoyl]oxy]methyl]-5,10,15,20-tetrakis(R,R,R,R-o-pivalamido)phenylporphyrinatoiron(II); Fe(II)TpivPP, 5,10,15,20-tetrakis(R,R,R,R-o-pivalamido)phenylporphyrinatoiron(II); HSA, human serum albumin; r-HSA, recombinant human serum albumin; HSA-Fe(II)P, Fe(II)P bound to HSA; Asb, L-ascorbic acid; DMF, N,N-dimethylformamide; DMSO, dimethyl sulfoxide; τ1/2, halflife of the oxygen adduct; P1/2(O2), oxygen affinity; OTE, oxygen transporting efficiency; BCG, bromocresol green; kon, rate constant of oxygen binding; koff, rate constant of oxygen dissociation; P1/2(CO), carbon monoxide affinity; 1,2-Me2Im, 1,2dimethylimidazole; SOD, superoxide dismutase; lipidheme, tetraphenylheme derivative having four alkylphosphocholine groups; LH-V, lipidheme vesicle; LH-M, lipidheme microsphere.

S1043-1802(97)00090-6 CCC: $14.00

Figure 1. Structure of Fe(II)P.

We have synthesized 2-[[[8-[N-(2-methylimidazolyl)]octanoyl]oxy]methyl]-5,10,15,20-tetrakis(R,R,R,R-o-pivalamido)phenylporphyrinatoiron(II) [Fe(II)P with a molecular mass of 1.3 kDa] which has an axial base covalently bound to the porphyrin ring (Figure 1). The axial base coordinates to the fifth coordination site of the central iron of the porphyrin and permits dioxygen coordination at the sixth site. We confirmed that Fe(II)P complexed with oxygen in organic solvents such as toluene or N,N-dimethylformamide (DMF) without adding imidazole derivatives for axial bases. Recently, we have found that the injection of a Fe(III)P/dimethyl sulfoxide (DMSO) solution into an HSA aqueous solution leads to a 1/1 complex of Fe(III)P and HSA (Komatsu et al., 1995), which gave a stable oxygen adduct even in the aqueous phase after its reduction by sodium dithionite (Na2S2O4) to Fe(II)P. Since then, the preparation procedure has been significantly improved to make more concentrated solutions of complex and to bind a greater number of Fe(II)P molecules into one HSA molecule so as to make the oxygen transporting ability comparable to that of hemoglobin. This paper reports on a new preparation procedure of HSA-Fe(II)P complexes binding one, four, and eight Fe(II)P molecules and also © 1997 American Chemical Society

Albumin−Porphyrinatoiron(II) Complexes

describes the characteristics and oxygen-binding properties of the resulting complexes. EXPERIMENTAL PROCEDURES

Materials. Fe(III)P was synthesized according to our previous paper (Tsuchida et al., 1995). Tetrapivalamidophenylporphyrinatoiron(III) [Fe(III)TpivPP, the socalled “picket-fence heme”] without an axial base was also synthesized according to the method reported by Collman et al. (1973). An HSA solution was purchased from Bayer Co., Ltd. (Albumin Cutter, 5 wt %), and the recombinant HSA solution (r-HSA, 25 wt %) was a gift from Green Cross Co., Ltd. Pure water (Otsuka Pharmaceutical Co., Ltd.) was used for dilution of HSA, and high-grade ethanol (purity, >99.5%; Kanto Chemical Co., Ltd.) was for the Fe(II)P solvent. Preparation of HSA-Fe(II)P. A brown-red Fe(III)P ethanol solution [[Fe(III)P] ) 0.15 mM] was prepared by dissolving Fe(III)P (39.1 mg) into 200 mL of ethanol. An HSA aqueous solution ([HSA] ) 18.8 µM) was prepared by diluting 10 mL of HSA (5 wt %) 40 times with pure water. L-Ascorbic acid (Asb, 104 mg) was completely dissolved into 2 mL of pure water under a N2 atmosphere, and 100 µL of the Asb solution was added to the Fe(III)P ethanol solution after bubbling CO through it for 10 min. The color of the Fe(III)P solution immediately changed to red, indicating reduction to Fe(II)P. A 200 mL aliquot of the Fe(II)P(CO) solution was dropped into 400 mL of the HSA solution and vigorously stirred, giving an Fe(II)P/HSA ratio of 4. The resulting red transparent mixture (600 mL) was evaporated down to 300 mL at 35 °C for 3 h and further concentrated to 7 mL by ultrafiltration (cutoff MW of 50 000, N2 pressure of 5 kg/cm2, 2 h, 4 °C). The remaining or oxidized Asb molecules were also removed during this procedure. The solution diluted 10 times was freeze-dried to obtain a reddish-white powder for long-term storage. The powder was redispersed by dissolving it in phosphate-buffered saline (300 mM, pH 7.4). After the dialysis against a saline solution for 15 h, the concentration of the sample was adjusted to 3 mM Fe(II)P for the complex where Fe(II)P/HSA ) 4. The HSA-Fe(II)P solution which bound eight Fe(II)P molecules to one HSA was also prepared using the same procedure. In this case, the HSA concentration was decreased to half ([HSA] ) 9.4 mM), and the Fe(II)P concentration after preparation was adjusted to 6 mM. Determination of the Half-Life of the Oxy State of HSA-Fe(II)P and Its O2 Affinity. One milliliter of the HSA-Fe(II)P(CO) solution in a 1 cm quartz cuvette was converted to the oxygenated state (oxy state) by irradiating with visible light from a halogen lamp (500 W) under flowing O2. The cuvette was constantly cooled in an ice bath. The conversion to the oxy state was completed within 6 min, and the oxy state identified in the change of the maximum absorption wavelength (λmax) from 540 to 548 nm (Q-band). The half-life of the oxy state (τ1/2) at 37 °C was determined from a decrease in the Q-band peak at 548 nm. The oxygen-binding and dissociation equilibrium curve was plotted from the visible spectral change after the complex was saturated with O2/N2 mixed gases in which the oxygen partial pressure decreased from 760 Torr (pure O2 gas) to zero. The 100% oxy state and the 100% deoxy state were determined from the spectra after the solution was saturated with 100% O2 gas and 100% N2 gas, respectively. The oxygen affinity was evaluated as P1/2(O2), the oxygen partial pressure where 50% of Fe(II)P was oxygenated. Quantitative Analysis of HSA and Fe(II)P. The HSA concentration in a HSA-Fe(II)P(CO) solution was determined using a bromocresol green method (BCG

Bioconjugate Chem., Vol. 8, No. 4, 1997 535

method, A/G-B test wako, Wako Pure Chemical Industries). The absorption at 630 nm was measured after reacting BCG with HSA. A calibration curve was obtained by diluting a standard HSA solution ([HSA] ) 1-5 g/dL). To avoid the influence of Fe(II)P(CO) absorption, the difference spectra between the HSA-Fe(II)P(CO)/BCG solution and HSA-Fe(II)P(CO)/pure water were recorded. After the injection of the ethanolic Fe(II)P(CO) solution into water, the suspended Fe(II)P(CO) can be completely extracted with chloroform because of its low solubility into water, whereas Fe(II)P(CO) in HSA dissolved in aqueous solutions cannot be extracted with chloroform. The amount of Fe(II)P bound to HSA can therefore be determined by extracting the Fe(II)P with a chloroform/ methanol (1/1) mixture. After the complete drying of the extracted solution, it was redissolved in methanol, and its concentration was determined from the molar absorption coefficient of Fe(II)P(CO) (417nm in methanol ) 2.0 × 105 M-1 cm-1). The binding ratio of Fe(II)P to HSA was determined from the quantitative analyses of Fe(II)P and HSA. Viscosity of HSA-Fe(II)P. The viscosity of the HSA-Fe(II)P solution was measured using a cone plate rotatory viscometer (Shibaura System VSA-K). Kinetic Parameters of O2 and CO Binding to HSA-Fe(II)P. The binding (kon) and dissociation (koff) rate constants of O2 and CO for HSA-Fe(II)P in aqueous solutions [[Fe(II)P] ) 20 µM] were determined by laser flash photolysis spectroscopy (Gibson, 1970; Traylor et al., 1985) using an UNISOKU TSP-601 spectrometer. CO affinity [P1/2(CO)] was determined from the visible absorption spectral change for various partial pressures of CO. RESULTS AND DISCUSSION

The oxygen-binding properties of Fe(II)P having an intramolecular axial base and Fe(II)TpivPP with no covalently bound axial base were compared in toluene solutions. Fe(III)TpivPP and Fe(III)P were reduced in toluene by aqueous Na2S2O4. When a large molar excess of the axial base [1,2-dimethylimidazole (1,2-Me2Im)] was coexisting, the Fe(II)TpivPP formed a stable oxygen adduct, which was confirmed by the visible absorption spectrum (λmax ) 421 and 544 nm). On the other hand, Fe(II)P gave an oxygen adduct without adding 1,2-Me2Im. The coordination of oxygen to the sixth coordination site of Fe(II)P was also confirmed from the visible absorption spectrum (λmax ) 422 and 549 nm). P1/2(O2) of Fe(II)P in toluene was 38 Torr at 25 °C. No difference in P1/2(O2) between Fe(II)P and Fe(II)TpivPP was observed in the toluene solution. After carbonylation under a 20-fold molar excess of 1,2-Me2Im in ethanol, the CO adduct of Fe(II)TpivPP was mixed with HSA, and then an oxygenated HSA-Fe(II)TpivPP complex was obtained by irradiating with visible light under an oxygen atmosphere. Nevertheless, the half-life of the oxygen adduct was only about 10 min at 37 °C. The coordination of the histidine residues of HSA to protoheme and deuteroheme was reported by Casella et al. (1993). The histidine residues of HSA do not however allow axial coordination to Fe(II)P, since complexation with histidine residues and oxygen binding was not observed for a Fe(II)TpivPP/HSA system. This result might be due to the binding of Fe(II)P to HSA by the hydrophobic interaction between the polypeptide chain and tert-butyl groups of Fe(II)P so that no histidine residue of HSA can coordinate to the central iron of Fe(II)P. Furthermore, the excess imidazole would cause unfavorable problems by binding to the site of the oxygen coordination, binding to HSA, or being toxic through intravenous injection. On the other hand, because Fe(II)P possesses only the intramolecular imi-

536 Bioconjugate Chem., Vol. 8, No. 4, 1997

Figure 2. Visible absorption spectral change after O2 and CO binding to HSA-Fe(II)P in an aqueous medium (pH 7.4 and 25 °C).

dazole, and carbonylation in the ethanol solution allows 100% coordination at the fifth coordinated site of Fe(II)P, Fe(II)P can be bound to HSA without changing the coordination structure. Figure 2 shows the visible absorption spectra of HSAFe(II)P in aqueous solutions. When a reducing agent such as Asb was added to Fe(III)P bound to HSA under a N2 atmosphere, only four-coordinated Fe(II)P (λmax ) 438 and 560 nm) was observed. In contrast, carbonylation was once carried out in an organic solution, the intramolecular coordination of the imidazole to the fifth coordinated site was confirmed. After the incorporation of Fe(II)P(CO) into HSA, five-coordinated deoxy-Fe(II)P (λmax ) 439, 542, 563, and 605 nm) was obtained by irradiating HSA-Fe(II)P(CO) with visible light under a N2 atmosphere. The oxy state (λmax ) 424 and 548 nm) was confirmed after passing oxygen gas over the solution of the deoxy state. This conversion is reversible, and the degree of oxygenation corresponds to the oxygen partial pressure. The carbonyl state (λmax ) 424 and 540 nm) was also immediately generated upon exposure to carbon monoxide gas with either the deoxy or oxy state. The molar absorption coefficients ( at λmax of the Soret band) of HSA-Fe(II)P were 1.1 × 105 M-1 cm-1 (deoxy state), 9.8 × 104 M-1 cm-1 (oxy state), and 1.3 × 105 M-1 cm-1 (carbonyl state). The solution of HSA-Fe(II)P having four Fe(II)P molecules was prepared by mixing an Fe(II)P ethanol solution (0.15 mM) with a diluted HSA aqueous solution (18.8 µM). The evaporation of ethanol, the concentration by ultrafiltration, and the dialysis could be carried out with no degradation of HSA and Fe(II)P. HSA-Fe(II)P binding eight Fe(II)P molecules was also prepared when the concentration of Fe(II)P was reduced to half during the sample preparation. The preparation procedure as described above has been improved in comparison with the former reported method (Komatsu et al., 1995) in the following points. (1) Asb and ethanol were used instead of Na2S2O4 and DMSO. These chemicals are relatively harmless. (2) The ferric state of the iron-porphyrin derivative can be reduced to the ferrous state by the addition of 1 molar equiv of Asb in ethanol. (3) HSA-Fe(II)P(CO) can be stored as a freeze-dried powder under long-term storage (2 years), and the concentration can be adjusted by redissolving it in a given amount of pure water. There was no difference in the turbidity, the Fe(II)P incorporation ratio, and the filter permeability for the HSA-Fe(II)P solution before

Tsuchida et al.

Figure 3. Relationship between mixing ratio and binding ratio of Fe(II)P/HSA. Table 1. Comparison of Solution Properties at 37 °C solutions

specific gravity

viscosity (cP at 230 s-1)

HSA-Fe(II)P serum human blood

1.013-1.021 1.027 1.055-1.063

1.3 1.3 4.4-5.0

and after redissolution of the freeze-dried powder. HSAFe(II)P(CO) could be stored at 4 °C for more than 1 month even in the solution state. The carbonyl state was easily converted to the oxy state by the irradiation of visible light under an oxygen atmosphere. The quantitative analyses of the free Fe(II)P after the preparation of HSA-Fe(II)P disclosed an incorporation ratio of 100% for Fe(II)P/HSA ) 1, 99% for 4, 94% for 8, but only 60% for 14 as shown in Figure 3. The concentration of HSA after sample preparation was 5.0 g/dL for Fe(II)P/HSA ) 1, 5.1 g/dL for 4, and 4.9 g/dL for 8, indicating no loss of HSA during the sample preparation. Figure 3 indicates a possible binding ratio of Fe(II)P to one HSA molecule of about eight. Interestingly, the molar absorption coefficients of the Fe(II)P(CO) in HSA were constant (540nm ) 1.2 × 104 M-1 cm-1 in phosphatebuffered saline at pH 7.4 and 25 °C) versus the Fe(II)P/ HSA ratio ()1, 4, and 8). The solution properties of HSA-Fe(II)P are summarized in Table 1. The specific gravity varies from 1.013 [Fe(II)P/HSA ) 0] to 1.021 [Fe(II)P/HSA ) 8]. The viscosity of HSA-Fe(II)P was the same as that of HSA (1.3 cP at a high share rate of 230 s-1), which was much lower than that of human blood (4.4-5.0 cP). There is no indication of aggregation even when eight Fe(II)P molecules are bound to one HSA, indicating its excellent dispersion stability with high oxygen transporting ability. Figure 4 shows the spectral change of the Q-band for the HSA-Fe(II)P(O2) incubated at 37 °C in air. The absorption maxima at 548 nm just before or after incubation of the oxygenated complex for up to 17 h gave Fe(III)P percentages of 0 or 100%, respectively. The halflife of the oxygen adduct (τ1/2) at 37 °C was about 1 h. This value was constant for HSA-Fe(II)P at different ratios (Table 2). A stable τ1/2 was obtained when an equimolar amount of Asb was added to Fe(III)P. The addition of superoxide dismutase and catalase did not improve the lifetime, indicating that there was no influence of active oxygen species generated from the oxidation of Fe(II)P(O2) on the half-life. A significantly short τ1/2 was observed when an excess amount of Asb was added to the Fe(III)P for reduction to Fe(II)P, and the effect of superoxide dismutase and catalase confirmed at

Bioconjugate Chem., Vol. 8, No. 4, 1997 537

Albumin−Porphyrinatoiron(II) Complexes

Figure 4. Oxidation of Fe(II)P(O2) bound to HSA. Fe(II)P/HSA ) 4 mol/mol. [HSA]: 5 wt %, under air, at pH 7.4 and 37 °C. Table 2. τ1/2 of Oxygenized HSA-Fe(II)P Prepared in Various Conditionsa samples HSA-Fe(II)P

blood

[HSA] [Fe(II)P] Fe(II)P/ O2 transportb τ1/2 (mM) (mM) HSA (mL/L) (min) 0.750 0.025 0.050 0.125 0.750 0.750 -

0.75 0.10 0.20 0.50 3.00 6.00 9.2c

1/1 4/1 4/1 4/1 4/1 8/1 4/1d

4.2 0.6 1.1 2.8 17.0 34.0 59

60 63 69 65 62 59 -

a In phosphate-buffered saline at pH 7.4 and 37 °C. b O 2 transport ) OTE (percent)/100 × [heme] (molar) × 22.4 L/mol × 310 K/273 K. c [heme]. d Heme/Hb.

Figure 5. O2 binding equilibrium curve of HSA-Fe(II)P (pH 7.4 and 37 °C): (b) Fe(II)P/HSA ) 4 mol/mol, (2) Fe(II)P/HSA ) 8 mol/mol.

the time indicated the generation of active oxygen species from the autoxidation of Asb, and their oxidation of Fe(II)P to Fe(III)P. Figure 5 shows the oxygen-binding and dissociation equilibrium curves of HSA-Fe(II)P [Fe(II)P/HSA ) 4] obtained from its saturated spectrum at each oxygen partial pressure. From this figure, the oxygen affinity [P1/2(O2)] was estimated to be about 30 Torr, and it was expected that HSA-Fe(II)P would release 22% of the

bound oxygen, if it circulated between the lung (PO2 ) 110 Torr) and mixed venouses (PO2 ) 40 Torr). The same results were obtained for HSA-Fe(II)P having eight Fe(II)P. The Hill coefficient was calculated to be 1.0 from the curve, and no allostericity was observed. The half-lives (τ1/2) of the Fe(II)P(O2) at 1, 4, 25, and 37 °C were 7 days, 5 days, 7 h, and 1 h, respectively. The activation energy was calculated as 89 kJ/mol. The τ1/2 was increased by raising the oxygen partial pressure (e.g., τ1/2 was 140 min at 37 °C in a pure oxygen atmosphere). This is the same profile as that of hemoglobin (Sakai et al., 1994). Deoxyhemoglobin, to which H2O is weakly coordinated, reacts with oxygen more rapidly to form methemoglobin than does oxyhemoglobin (Brantley et al., 1993). The O2 and CO binding parameters are summarized in Table 3. The P1/2(O2) of HSA-Fe(II)P in an aqueous solution is 8.0 Torr at 25 °C and 30 Torr at 37 °C. The P1/2(O2) of Fe(II)P in toluene is 38 Torr at 25 °C. The low P1/2(O2) of Fe(II)P in HSA means that the oxygen affinity is higher than that of Fe(II)P in the solution state. This might be due to the existence of polar amide groups of HSA around the oxygen coordination site of Fe(II)P (Springer et al., 1989). Furthermore, kon and koff are extremely high in comparison with those of red blood cells (1.1 × 104 M-1 s-1, 0.16 s-1) and hemoglobin (3.3 × 107 M-1 s-1, 13 s-1). Because no difference in the τ1/2 of HSA-Fe(II)P with different binding ratios of Fe(II)P was observed, hydrophobicity should originate from the tert-butyl groups of Fe(II)P, not from the hydrophobic pocket of HSA. Therefore, we expect to improve the half-life of the oxygen adduct as an oxygen carrier by designing the molecular structure of the heme derivative to be more hydrophobic such as a double-sided porphyrin which has four alkyl chains on both sides of the porphyrin plane (Tsuchida et al., 1993). We have been developing artificial red cells using totally synthetic heme derivatives (lipidheme) (Kobayashi & Tsuchida, 1995). The heme derivatives were amphiphiles and were incorporated into the hydrophobic region of the bilayer of phospholipid vesicles (LH-V) or onto the surface of a triglyceride microsphere (LH-M). The P1/2(O2) values of these oxygen carriers at 37 °C were 43 and 38 Torr, respectively. These are higher than that of red blood cells (27 Torr). Their τ1/2 values at 37 °C were 24 and 12 h, respectively. In the case of Fe(II)P, it could not form a stable oxygen adduct in the phospholipid bilayer membrane because it does not have alkylphosphocholine groups which provide a high affinity with the phospholipid bilayer membrane and fix the iron-porphyrin at the hydrophobic center of the bilayer membrane. However, Fe(II)P can easily bind to HSA, and the advantages of HSA are considered to be its low toxicity and long circulation lifetime in comparison with the others. The rate constants of oxygen binding (kon) of LH-V and LH-M were 9.8 × 107 and 9.6 × 107 M-1 s-1, respectively.

Table 3. O2 and CO Binding Parameters of the HSA-Fe(II)P in an Aqueous Medium at 25 °C O2 samples HSA-Fe(II)P r-HSA-Fe(II)Pa Fe(II)Pb Hb (R state)c red celld

solutions pbd

(pH 7.4) pbd (pH 7.4) toluene pbd (pH 7.0) pbd (pH 7.4)

P1/2 (Torr) (30)f

8.0 11 (33)f 38 0.22 (0.82)f 8.8 (27)f

CO

10-7kon (M-1 s-1)

10-3koff (s-1)

102P1/2 (Torr)

10-6kon (M-1 s-1)

102koff (s-1)

24 23 16 3.3 0.0011

3.2 4.2 46 0.013 0.00016

1.4 1.7 0.6 0.14 57

4.4 4.4 2.9 4.6 0.014

8 9 17 0.9 1

a r-HSA, recombinant HSA. b Measured by Tsuchida (1995). c Measured by Gibson (1970). d Measured by Tsuchida (1985). e pb, phosphate buffer. f P1/2 values in parentheses were measured at 37 °C.

538 Bioconjugate Chem., Vol. 8, No. 4, 1997

kon and koff are quite high relative to those of hemoglobin, indicating its fast oxygen binding and dissociation. This would be due to the easily accessible pathway to the central iron of lipidheme through four alkylphosphocholine groups rather than through the folded polypeptide main chain of hemoglobin. In the case of red blood cells, the kinetic parameters are apparent and very slow because of the small surface area for oxygen transport and low diffusion of the viscous and concentrated hemoglobin solution inside the cell. In the case of HSA-Fe(II)P, kon is significantly high (2.4 × 108 M-1 s-1) compared with that of hemoglobin. There is no steric hindrance of the peptide chain around the oxygen binding sites like a heme pocket of hemoglobin, but the environment around them is kept hydrophobic. The koff of HSA-Fe(II)P is 3.2 × 103 s-1, which is about 1/ that of LH-V (8.2 × 103 s-1) and LH-M (1.0 × 104 s-1). 3 This would be due to the vicinity of the oxygen binding site of Fe(II)P having a higher polarity than the hydrophobic region of LH-V and LH-M. This was supported by the result that Fe(II)P in toluene showed a koff (4.6 × 104 s-1) 10 times higher in comparison with that in HSA. It is quite remarkable that eight Fe(II)P molecules can be bound to HSA. The hydrophobic interaction of the four tert-butyl groups of Fe(II)P with the hydrophobic region of the HSA surface enables the binding of Fe(II)P onto the surface. Because all Fe(II)P bound onto the surface would have the same chemical environment that comes from the hydrophobicity of the four tert-butyl groups, the molar absorption coefficient, the binding parameters of oxygen and carbon monoxide, and the half-life of the oxygen adduct do not depend on the binding ratio of Fe(II)P. From the oxygen transporting efficiency (22%) in Figure 5, the transporting amount of oxygen was calculated as shown in Table 2. The concentration of HSA in an HSA-Fe(II)P solution was adjusted to that in blood (5 wt %; 0.75 mM) to control the colloidal osmotic pressure of the solution. When a 5 wt % HSA-Fe(II)P [Fe(II)P/HSA ) 8] solution is prepared, it can transport 3.4 mL/dL of oxygen during the circulation between lung (PO2 ) 110 Torr) and mixed venouses (PO2 ) 40 Torr). This corresponds to about 60% of the oxygen transporting amount of human blood (5.9 mL/dL), because the heme concentration of the HSA-Fe(II)P solution is low (6.0 mM) in comparison with that (9.2 mM) of blood ([hemoglobin] ) 15 g/dL). Furthermore, the value of OTE can be increased by using other heme derivatives having a lower P1/2(O2) than that of Fe(II)P, and the transporting amount of oxygen can be improved by increasing the HSA concentration from 5 wt % for in vivo use. Recently, recombinant HSA (r-HSA) has been manufactured by gene cloning and expression in Saccharomyces cerevisiae or Escherichia coli, etc. (Sumi et al., 1993). The same results were obtained using r-HSA-Fe(II)P as a totally synthetic oxygen carrier as shown in Table 3. Animal tests are now being undertaken by the authors. ACKNOWLEDGMENT

This work was partially supported by Grants-in-Aid from the Ministry of Education, Science, and Culture, Japan (05403028 and 07508005) and a Waseda University Grant for Special Research Projects. LITERATURE CITED Adams, P. A., and Berman, M. C. (1980) Kinetics and Mechanism of the Interaction between Human Serum Albumin and Monomeric Haemin. Biochem. J. 191, 95-102. Beaven, G. H., Chen, S.-H., D’Albis, A., and Gratzer, W. B. (1974) A Spectroscopic Study of the Haemin-Human-SerumAlbumin System. Eur. J. Biochem. 41, 539-546.

Tsuchida et al. Bonaventura, J., Brouwer, M., Brouwer, T., Cashon, B., and Cameron, S. (1994) presented in part at the 11th Congress ISABI, Boston, MA. Brantley, R. E., Jr., Smerdon, S. J., Wilkinson, A. J., Singleton, E. W., and Olson, J. S. (1993) The Mechanism of Autooxidation of Myoglobin. J. Biol. Chem. 268, 6995-7010. Casella, L., Gullotti, M., Poli, S., and Gioia, L. D. (1993) HaemProtein Interactions. The Binding of Haem Complexes to Serum Albumin. Gazz. Chim. Ital. 123, 149-154. Collman, J. P., Gagne, R. R., Halbert, T. R., Marchon, J. C., and Reed, C. A. (1973) Reversible Oxygen Adduct Formation in Ferrous Complexes Derived from a Picket Fence Porphyrin. Model for Oxymyoglobin. J. Am. Chem. Soc. 95, 7868-7870. Gibson, Q. H. (1970) The Reaction of Oxygen with Hemoglobin and the Kinetic Basis of the Effect of Salt on Binding of Oxygen. J. Biol. Chem. 245, 3285-3288. Hrkal, Z., and Klementova, S. (1984) Bilirubin and Haeme Binding to Human Serum Albumin Studied by Spectroscopy Methods. J. Biochem. 16 (7), 799-804. Kobayashi, K., and Tsuchida, E. (1995) Totally Synthetic Hemes: Their Characteristics and Oxygen-Carrying Capacity in Dogs. In Artificial Red Cells (E. Tsuchida, Ed.) pp 93116, John Wiley and Sons, New York. Kodicek, M., Hrkal, Z., and Vodrazka, Z. (1983) Kinetics of Haem Binding to Human Albumin and Haemopexin: Nonspecific Interactions of Haem with Proteins. Int. J. Biol. Macromol. 5, 194-198. Komatsu, T., Ando, K., Kawai, N., Nishide, H., and Tsuchida, E. (1995) O2-Transport Albumin: A New Hybrid-Haemoprotein Incorporating Tetraphenylporphyrinatoiron(II) Derivative. Chem. Lett., 813-814. Lamola, A. A., Asher, I., Muller-Eberhard, U., and Poh-Fitzpatrick, M. (1981) Fluorimetric Study of Binding of Protoporphyrin to Haemopexin and Albumin. Biochem. J. 196, 693-698. Marden, M. C., Hazard, E. S., Leclerc, L., and Gibson, Q. H. (1989) Flash Photolysis of the Serum Albumin-Heme-CO Complex. Biochemistry 28, 4422-4426. Moehring, G. A., Chu, A. H., Kulansik, L., and Williams, T. J. (1983) Heterogeneity of Albumin As Detected by Its Interactions with Deuteroporphyrin IX. Biochemistry 22, 3381-3386. Morgan, W. T., Smith, A., and Koskelo, P. (1980) The Interaction of Human Serum Albumin and Hemopexin with Porphyrins. Biochim. Biophys. Acta 624, 271-285. Sakai, H., Takeoka, S., Seino, Y., and Tsuchida, E.(1994) Suppression of Methemoglobin Formation by Glutathione in a Concentrated Hemoglobin Solution and in a HemoglobinVesicle. Bull. Chem. Soc. Jpn. 67, 1120-1125. Springer, B. A., Egeberg, K. D., and Sligar, S. G. (1989) Discrimination between Oxygen and Carbon Monoxide and Inhibition of Autooxidation by Myoglobin. J. Biol. Chem. 264, 3057-3060. Sumi, A., Ohtani, W., Kobayashi, K., Ohmura, T., Yokoyama, K., Nishida, M., and Suyama, T. (1993) Purification and Physicochemical Properties of Recombinant Human Serum Albumin. Colloq. INSERM 227, 293-298. Traylor, T. G., Tsuchiya, S., Campbell, D., Mitchell, M., Stynes, D., and Koga, N. (1985) Anthracene Heme Cyclophanes: Steric Effect in CO, O2, and RNC Binding. J. Am. Chem. Soc. 107, 604-614. Tsuchida, E., Nishide, H., Yuasa, M., Hasegawa, E., Matsushita, Y., and Eshima, K. (1985) Phosphocholine-substituted 5,10,15,20-Tetraphenylporphyrinatoiron(II): Oxygen Carrier under Physiological Conditions. J. Chem. Soc., Dalton Trans., 275-278. Tsuchida, E., Komatsu, T., Arai, K., and Nishide, H. (1993) Synthesis and Dioxygen-binding Properties of Double-sided Porphyrinatoiron(II) Complexes bearing Covalently Bound Axial Imidazole. J. Chem. Soc., Dalton Trans., 2465-2469. Tsuchida, E., Komatsu, T., Kumamoto, S., Ando, K., and Nishide, H. (1995) Synthesis and O2-binding Properties of Tetraphenylporphyrinato-iron(II) Derivatives Bearing a Proximal Imidazole Covalently Bound at the β-pyrrolic Position. J. Chem. Soc., Perkin Trans. 2, 747-753.

BC9700906