Extension Arm Facilitated Pegylation of αα-Hemoglobin with

Oct 20, 2009 - ... load: https://cdn.mathjax.org/mathjax/contrib/a11y/accessibility-menu.js .... The structural and functional advantages of reversibl...
2 downloads 0 Views 1MB Size
Download PDF
2062

Bioconjugate Chem. 2009, 20, 2062–2070

Extension Arm Facilitated Pegylation of rr-Hemoglobin with Modifications Targeted Exclusively to Amino Groups: Functional and Structural Advantages of Free Cys-93(β) in the PEG-Hb Adduct Dongxia Li,†,‡ Tao Hu,†,‡ Belur N. Manjula,‡ and Seetharama A. Acharya*,‡,§ Departments of Physiology and Biophysics, and of Medicine, Albert Einstein College of Medicine, Bronx, New York 10461. Received April 14, 2009; Revised Manuscript Received August 30, 2009

Cys-93(β) of hemoglobin (Hb) was reversibly protected as a mixed disulfide with thiopyridine during extension arm facilitated (EAF) PEGylation and its influence on the structural and functional properties of the EAF-PEGHb has been investigated. Avoiding PEGylation of Cys-93(β) in the EAF-PEG-Hb lowers the level of perturbation of heme pocket, R1β2 interface, autoxidation, heme loss, and the O2 affinity, as compared to the EAF-PEG-Hb with PEGylation of Cys-93(β).The structural and functional advantages of reversible protection of Cys-93(β) during EAF PEGylation of oxy-Hb has been compared with Euro PEG-Hb generated by EAF PEGylation of deoxy Hb where Cys-93(β) is free in the final product. The RR-fumaryl cross-linking and EAF PEGylation targeted exclusively to Lys residues has been combined together for generation of second-generation EAF-PEG-Hb with lower oxygen affinity. The PEG chains engineered on Lys as well as PEGylation of Cys-93(β) independently contribute to the stabilization of oxy conformation of Hb and hence increase the oxygen affinity of Hb. However, oxygen affinity of the EAF-PEG-RR-Hb is more sensitive to the presence of PEGylation on Cys-93(β) than that of the EAF-PEG-Hb. The present modified EAF PEGylation platform is expected to facilitate the design of novel versions of the EAF-PEG-Hbs that can now integrate the advantages of avoiding PEGylation of Cys-93(β).

INTRODUCTION 1

PEGylation, conjugation with polyethylene glycol (PEG ) chains, can improve the therapeutic potentials of peptides and proteins by camouflaging the immune system and diminishing the in ViVo glomerular filtration (1-3). Recently, PEGylation of hemoglobin (Hb) has emerged as a novel chemical approach to neutralize the vasoactivity of acellular Hb, a major obstacle for its therapeutic application as Hb based oxygen carriers (HBOC) (4, 5). The PEGylated bovine Hb developed by Enzon Inc. (6) is one of the early PEG-Hbs and carries ten copies of PEG-5K chain per tetramer on amino groups (7). Another PEGylation protocol, extension arm facilitated (EAF) PEGylation (8-11), has been developed by us recently. The 2-iminothiolane (IT), an internally cyclized form of δ-mercapto butyrimidate, is used to introduce δ-mercapto butirimidyl chain * Address correspondence to Seetharama A. Acharya, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, Tel: 1-718-430-2133; Fax: 1-718-430-8819; E-mail: acharya@aecom. yu.edu. † The authors make same contribution on the work. ‡ Department of Physiology and Biophysics. § Department of Medicine. 1 ABBREVIATIONS: Hb, hemoglobin; PEG, polyethylene glycol; PEGylation, conjugation with polyethylene glycol; EAF, extension arm facilitated; PEG-Hb, the PEGylated hemoglobin; RR-Hb, RR-fumaryl cross-linked hemoglobin; PEG-RR-Hb, the PEGylated RR-Hb; PBS, phosphate buffered saline; TFA, trifluoroacetic acid; Kd, the dissociation constant; HBOC, hemoglobin based oxygen carrier; P50, partial oxygen pressure at half saturation; SEC, size exclusion chromatography; RPHPLC, reverse phase high performance liquid chromatography; CD, circular dichroism; n, Hill coefficient; 2,3-DPG, 2,3-diphosphoglycerate; 4-PDS, 4,4′-dithiopyridine; IT, 2-iminothiolane; TCEP, Tris(2-carboxyethyl) phosphine; NEM, N-ethyl maleimide; kox, the first-order autoxidation rate constant; HSA, human serum albumin; kfast, the rate constant for the fast phase of heme loss.

on the ε-amino groups of Hb. The newly engineered thiol group is then PEGylated using PEG maleimide. The hexaPEGylated Hb, (SP-PEG5K)6-Hb generated by the EAF PEGylation, has served as the model molecule to develop MP4 (a hexa-PEG-Hb at 4 g/dL) that is in phase III clinical trial. (SP-PEG5K)6-Hb carries four PEG-5K chains on the extrinsic thiol groups engineered on the ε-amino groups and two on the intrinsic thiols of two Cys-93(β) residues of Hb. Thus, EAF PEGylation of oxy Hb introduces two PEG-5K chains by direct PEGylation of Cys-93(β) (i.e., without an extension arm). Cys-93(β) is located ∼1.4 nm from the heme and is adjacent to proximal His-92(β) that coordinates to the center iron of a heme (12). Proton NMR spectra has shown that chemical modification of Cys-93(β) significantly influences the heme pocket of the β chain (13). (SP-PEG5K)2-Hb, a Hb derivative with PEGylation exclusively on Cys-93(β), shows a change in the Fe-proximal His residue stretching frequency (14). Manjula et al. (15) reported that (SP-PEG5K)2-Hb exhibits a higher oxygen affinity and a lower Hill coefficient than Hb. PEGylation of Cys-93(β) also enhanced the autoxidation rate and the rate of heme loss from metHb (16). Thus, PEGylation of Cys-93(β) unfavorably influences the structural and functional properties of Hb. These unfavorable structural and functional changes could be minimized or eliminated in the EAF-PEGHb by avoiding PEGylation of Cys-93(β). The high oxygen affinity of (SP-PEG5K)6-Hb was one of the attractive features for Sangart in developing MP4. The high oxygen affinity is expected to achieve a targeted O2 delivery in ViVo to the hypoxic regions (7). The high O2 affinity is also expected to minimize the O2 unloading on the arterial side of circulation and thus reduce the possible vasoactivity that is induced through antoregulatory mechanism. Though high O2 affinity of the EAF-PEG-Hb has been attributed to the PEGylation of Cys-93(β) initially (15), the EAF PEGylation of rHb[Cys-93(β)fAla] suggested that the high O2 affinity of the

10.1021/bc900170e CCC: $40.75  2009 American Chemical Society Published on Web 10/20/2009

Second Generation of EAF-PEG-Hbs

Bioconjugate Chem., Vol. 20, No. 11, 2009 2063

EXPERIMENTAL PROCEDURES

Figure 1. Schematic representation of the extension arm facilitated PEGylation of Hb with reversible protection of Cys-93(β).

EAF-PEG-Hb is an intrinsic aspect of the PEG-Hb and is presumably a consequence of the hydrated PEG-shell around Hb (10). Recent studies have suggested that the O2 affinity of the PEGHb may have to be lower than that of MP4 to achieve an efficient tissue oxygenation (17). The high oxygen affinity of the EAF-PEG-Hb represents the synergistic effects of EAF PEGylation at the amino groups and the direct PEGylation at Cys-93(β). Avoiding the direct PEGylation of Cys-93(β) may be necessary to generate the PEG-Hb with an O2 affinity lower than the ones in the current version of the EAF-PEG-Hb. EAF PEGylation of deoxy Hb is an approach that can protect Cys93(β) from PEGylation (9, 36). Reversible protection of Cys93(β) during EAF PEGylation of oxy-Hb is an alternative approach developed in the present study. EAF PEGylation platform has been modified here by designing reversible protection of Cys-93(β). Briefly, the thiol group of Cys-93(β) of oxy-Hb is converted to a mixed disulfide form by reaction with 4,4′-dithiopyridine (4-PDS), followed by EAF PEGylation of the modified Hb and the release of the thiopyridyl group using Tris(2-carboxy-ethyl) phosphine (TCEP) (Figure 1). A low oxygen affinity Hb, RR-fumaryl cross-linked Hb (RRHb), was also EAF PEGylated with protection of Cys-93(β) in order to generate a PEG-Hb with lower oxygen affinity. The influence of avoiding the PEGylation of Cys-93(β) on the generation of a doubly modified Hb (RR-cross-linking and EAFPEGylation) has also been investigated. This modified EAF PEGylation platform is expected to integrate the structural and functional advantages of protection of Cys-93(β) in the EAFPEG-Hbs.

Preparation of the PEG-Hbs. Human adult hemoglobin (HbA) was purified from human erythrocytes as previously described (18). RR-Hb was prepared as previously described (19). The conditions for EAF PEGylation were slightly modified from Manjula et al. (9). Briefly, HbA or RR-Hb at 1.0 mM (tetramer) was reacted with 10 mM maleimide phenyl-PEG5K (Malphe-PEG5K, Bioaffinity Inc., Rockford, IL) or 10 mM N-ethyl maleimide (NEM) in the presence of 5 mM IT in PBS buffer (pH 7.4) at 4 °C for 4 h. The resultant products are dialyzed extensively against PBS buffer (pH 7.4). For removal of free PEG, the sample was centrifuged at 6000 g using Centricon with 50 K cutoff membrane (Millipore) five times. The retentate was stored at -80 °C. Preparation of the PEG-Hbs with Reversible Protection of Cys-93(β). HbA or RR-Hb at 0.25 mM (tetramer) was reacted with 5 mM 4-PDS in PBS buffer (pH 7.4) at 4 °C for 1 h. The resultant samples were dialyzed extensively against PBS buffer (pH 7.4). The absence of 4-PDS was confirmed by adding excess TCEP that did not change the absorbance at 324 nm as compared to the control. The concentrated modified Hb (1 mM as tetramer) was reacted with 10 mM Malphe-PEG5K or 10 mM NEM in the presence of 5 mM IT in PBS buffer (pH 7.4) at 4 °C for 4 h. NEM was added to terminate the reaction. The resultant samples were centrifuged at 6000 g using Centricon with 50 K cutoff membrane five times to remove excess free PEG. A large excess of TCEP is added to release the thiopyridyl moiety from the disulfide moiety. For removal of TCEP, the sample was centrifuged at 6000 g using Centricon with 50 K cutoff membrane five times. The retentate was stored at -80 °C. Analytical Methods. Size exclusion chromatography (SEC) analysis was carried out on two Superose 12 columns (GE Healthcare, 1 × 31 cm2) in series at room temperature. The columns were equilibrated and eluted with PBS buffer (pH 7.4) at a flow rate of 0.5 mL/min. Reverse-phase HPLC analysis was carried out on a Vydac C4 column (4.6 × 250 mm2). The sample was eluted using a linear gradient of 35-50% acetonitrile containing 0.1% TFA in 100 min and then 50-70% acetonitrile containing 0.1% trifluoroacetic acid (TFA) in 30 min at a flow rate of 1.0 mL/min. SDS-PAGE analysis was carried out on a precast 14% tris-glycine gel (Invitrogen Inc.). Molecular radius of the Hb samples (1 mg/mL) was measured using a dynamic light scattering instrument (DynaPro, Protein Solutions, Lakewood, NJ). Autoxidation Experiments. Autoxidation studies of the PEG-Hb are carried out by incubating the Hb samples at 25 µM (1.6 mg/mL in tetramers) in PBS buffer (pH 7.4) plus 0.3 mM EDTA in sealed tubes at 37 °C. Aliquots of the sample were taken out at various time intervals. The first-order autoxidation rate constant (kox) of the Hb sample was measured and calculated as in the method by Hu et al. (16). The PEGHbs (40 mg/mL) in PBS (pH 7.4) were mixed with the bovine plasma (Sigma) at a volume ratio of 1:1. These diluted samples were used to simulate 50% exchange transfusion of the blood volume. The percentage of plasma metHb was measured as previously described (16). Heme Exchange Experiment. Heme exchange rate between metHb and human serum albumin (HSA) (Sigma) was measured as previously described (20, 21). MetHb was formed by reaction of oxyHb with potassium ferricyanide (22). The ferro- and ferricyanide were removed by passing the reaction mixture through a Sephadex G-25 column in 50 mM BisTris-Ac buffer (pH 7.5) containing 0.1 M NaCl. The heme exchange reaction was measured as in the method of Hu et al. (16). Tetramer-Dimer Dissociation Constant. Tetramer-dimer dissociation constant (Kd) of the PEG-Hbs was measured essentially as Manning et al. (23). Briefly, Hb samples in PBS

2064 Bioconjugate Chem., Vol. 20, No. 11, 2009

buffer (pH 7.4) were diluted to a series of Hb concentrations and subjected to two Superose 12 columns (1 × 31 cm2) in series. The columns were equilibrated and eluted with PBS buffer (pH 7.4) at the flow rate of 0.5 mL/min. RR-Hb was set as the tetramers for HbA. The PEG-RR-Hbs were set as the tetramers for the PEG-Hbs. The Hb samples in the presence of PBS buffer containing 0.9 M MgCl2 were set as the dimers. Circular Dichroism Spectroscopy. Circular dichroism (CD) spectra of the Hb samples were recorded on a JASCO-J815 spectropolarimeter (JASCO, Tokyo, Japan) at 25 °C with a 0.2 cm light path cuvette (310 µL). For the CD spectra from 250 to 200 nm, the Hb concentration was 1.3 µM as tetramer. For the CD spectra from 480 to 250 nm, the Hb concentration was 26.0 µM as tetramer. All the Hb samples were in PBS buffer, pH 7.4. The molar ellipticity (θ) is expressed in deg.cm2/dmol on a heme basis. Front-Face Fluorescence Measurements. Intrinsic frontface fluorescence measurements of the Hb samples were performed using a Shimadzu RF-5301 spectrofluorimeter at room temperature. The emission spectra were recorded from 300 to 350 nm using an excitation wavelength of 280 nm. Excitation and emission slit widths were both 5 nm. All the samples used were at Hb concentration of 6.0 mg/mL in PBS, pH 7.4. Cuvette with 1 cm path-length was used. Oxygen Affinity Measurements. Oxygen equilibrium measurements of the Hb samples were carried out using Hem-OScan (Aminco) at 37 °C in PBS buffer (pH 7.4). Triplicate measurements for each sample were performed to obtain the averaged values of P50 (oxygen affinity) and Hill coefficient (n). The Hb samples were at Hb concentration of 0.5 mM as tetramer in the presence and the absence of 2.5 mM 2,3-DPG.

Li et al.

Figure 2. Size exclusion chromatography analysis of the PEG-Hb (A) and the PEG-RR-Hb (B). (a) HbA; (b) the bisthiopyridyl Hb; (c) the PEG-Hb without protection of Cys-93(β); (d) the PEGylated bisthiopyridyl Hb; (e) the PEG-Hb with free Cys-93(β); (f) RR-Hb; (g) the bisthiopyridyl RR-Hb; (h) the PEG-RR-Hb; (i) the PEGylated bisthiopyridyl RR-Hb; (j) the PEG-RR-Hb with free Cys-93(β). All samples were loaded on two HR10/30 Superose 12 columns at room temperature. The column was eluted with PBS, pH 7.4, at 0.5 mL/min.

RESULTS EAF PEGylation of Un-Cross-Linked Hb with Protection of Cys-93(β). Size Exclusion Chromatographic Analysis of the PEG-Hbs. The SEC pattern of the PEG-Hbs is presented in Figure 2A. The elution position of the PEG-Hb (curve c) is earlier than that of HbA (curve a). The elution pattern of the PEGylated bisthiopyridyl Hb (curve d) is slightly right-shifted and exhibits some asymmetry on the descending side of the peak. This is presumably due to the absence of PEGylation on Cys-93(β). Moreover, the elution pattern shows a small peak at the position of Hb, reflecting the presence of minor unPEGylated thiopyridyl Hb. The elution pattern shows no noticeable changes upon releasing the thiopyridyl group from Cys-93(β) (curve e). Characterization of the PEG-Hb with Protection of Cys-93(β). RP-HPLC Analysis of the PEG-Hbs. RP-HPLC analysis of the PEG-Hbs is shown in Figure 3A. The HPLC pattern of Hb shows two peaks corresponding to β and R globin chains (curve a). Upon EAF PEGylation, the two globin chains decrease significantly, along with generation of one major and two minor elution peaks (curve b). These new peaks were estimated to correspond to the mono-, di-, and tri-PEGylated globin chains, according to the procedure by Hu et al. (24). The preferential PEGylation at β chain is apparently due to the efficient PEGylation at Cys-93(β). Since the absorbance of the PEG chains at 210 nm is very weak, the absorbance at 210 nm essentially detects the globin chains. Accordingly, the EAFPEG-Hb has been calculated to carry ∼5.7 PEG-5K chains per tetramer. The pattern of the PEGylated bisthiopyridyl Hb (curve c) shows minor thiopyridyl β-globin and unmodified R-globin, along with three new peaks. The peak for the monoPEGylated globin is asymmetrical and becomes symmetrical upon releasing the thiopyridyl moiety (curve d). This PEG-Hb was calculated to carry ∼5.0 PEG-5K chains per tetramer. It should be mentioned that the sum of peak areas in the HPLC pattern of

Figure 3. RP-HPLC analysis of the PEG-Hb (A) and the PEG-RR-Hb (B). (a) HbA; (b) the PEG-Hb; (c) the PEGylated bisthiopyridyl Hb; (d) the PEG-Hb with free Cys-93(β); (e) RR-Hb; (f) the PEG-RR-Hb; (g) the PEGylated bisthiopyridyl RR-Hb; (h) the PEG-RR-Hb with free Cys-93(β). The samples were loaded on a Vydac C4 column (0.46 × 25 cm2). The column was eluted with a linear gradient of 35-50% acetonitrile containing 0.1% TFA in 100 min, followed by 50-70% acetonitrile containing 0.1% TFA in 30 min at 1.0 mL/min.

the individual samples are taken as 100% for calculating the level of PEGylation in the respective PEG-Hbs. SDS-PAGE Analysis of the PEG-Hbs. The SDS-PAGE pattern of HbA shows a doublet band corresponding to R and β globin

Second Generation of EAF-PEG-Hbs

Bioconjugate Chem., Vol. 20, No. 11, 2009 2065

Figure 4. SDS-PAGE analysis of the PEG-Hbs and the PEG-RR-Hb. SDS-PAGE was carried out on a precast 14% Tris-glycine gel from the Invitrogen Corporation. Lanes 1 and 8, the molecular weight markers; lane 2, HbA; lane 3, the PEG-Hb; lane 4, the PEG-Hb with free Cys-93(β); lane 5, RR-Hb; lane 6, the PEG-RR-Hb; lane 7, the PEG-RR-Hb with free Cys-93(β). Proteins were identified by Coomassie blue staining. The PEG chain was detected by iodine staining. Table 1. Molecular Radius of the PEG-Hb Samples sample

radius (nm)

molecular volume (nm3)

HbA EAF-PEG-Hb without free Cys-93(β) EAF PEG-Hb with free Cys-93(β) RR-Hb EAF-PEG-RR-Hb without free Cys-93(β) EAF-PEG-RR-Hb with free Cys-93(β)

3.04 5.23 4.94 3.10 5.52 5.15

117.6 598.9 504.7 124.8 704.2 571.9

(∼16 kDa) (lane 2, Figure 4). EAF PEGylation of HbA resulted in the appearance of four bands. Iodine staining confirms that these bands (lane 3) correspond to one globin conjugated with 1, 2, 3, and 4 PEG-5K chains. When Cys-93(β) is reversibly protected, these bands are essentially unchanged (lane 4) except that the band corresponding to the tetraPEGylated globin is noticeably reduced. Molecular Radius of the PEG-Hbs by Dynamic Light Scattering. The molecular radius of Hb increased from 3.04 to 5.23 nm upon EAF PEGylation (Table 1). The molecular radius of the PEG-Hb decreased to 4.94 nm upon reversible protection of Cys-93(β), consistent with the results from the SEC and HPLC analysis. ReactiVity of Cys-93(β) of the PEG-Hbs. As reflected by the titration with 4-PDS (25), the PEG-Hb with protection of Cys93(β) has two free thiols per tetramer (curve b, Figure 5) and its reactivity to 4-PDS is comparable to that of HbA (curve a). Since the change in the reactivity of Cys-93(β) in the oxy state can be considered as an indicator of a change at the R1β2 interface in a chemically modified Hb, EAF PEGylation of Hb with protection of Cys-93(β) did not induce any perturbation at the R1β2 interface of Hb. In contrast, the reactivity of Cys93(β) of two other hexaPEG-Hbs, (Propyl-PEG5K)6-Hb (curve c), and (Propionyl-PEG5K)6-Hb (curve d) is significantly increased, indicating the structural perturbation at the R1β2 interface of the two PEG-Hbs. It should be mentioned that (Propyl-PEG5K)6-Hb and (Propionyl-PEG5K)6-Hb are prepared using reductive alkylation and acylation chemistry, respectively, where the PEG chains are directly conjugated to the amine groups of Hb (26, 27). Tetramer Stability of the PEG-Hbs. The Kd of HbA increased slightly upon EAF PEGylation with reversible protection of Cys93(β) (Figure 6). In contrast, the Kd of the EAF-PEG-Hb increased significantly. However, the Kd of HbA increased slightly upon NEM modification and further marginally de-

Figure 5. The reactivity of Cys-93(β) of the PEG-Hb (A) and the PEGRR-Hb (B). HbA (a), the PEG-Hb with free Cys-93(β) (b), (PropylPEG5K)6-Hb (c), (Propionyl-PEG5K)6-Hb (d), RR-Hb (e), the PEGRR-Hb with free Cys-93(β) (f), (Propyl-PEG5K)6-RR-Hb (g), and (Propionyl-PEG5K)6-RR-Hb (h) at the Hb concentration of 5 µM were incubated with 50 µM 4-PDS at room temperature in PBS, pH 7.4. The SH groups were estimated by measuring the conversion of 4-PDS to 4-thiopyridine at 324 nm as a function of time.

creased upon reversible protection of Cys-93(β). This indicated that the NEM modification at Cys-93(β) has little effect on the Kd of Hb. Accordingly, the PEG chains conjugated at Cys-93(β) play a predominant role in the tetramer stability of the EAFPEG-Hb. In contrast, the reductive alkylation and acylation chemistry based PEG-Hbs exist essentially as Hb dimers (26, 27). Structure and Function of the PEG-Hbs. Circular Dichroism Study of the PEG-Hbs. The L band (centered around 260 nm) is considered to be sensitive to the interactions between the heme and the surrounding globin, being influenced by the ligand interactions (28). As shown in the near-UV CD region,

2066 Bioconjugate Chem., Vol. 20, No. 11, 2009

Li et al.

Figure 6. The tetramer-dimer dissociation constant of the PEG-Hb. HbA (a), the PEG-Hb with free Cys-93(β) (b), the PEG-Hb without free Cys-93(β) (c), the NEM modified thiolated Hb with free Cys93(β) (d), and the NEM modified thiolated Hb without free Cys-93(β) (e) were loaded on the Superose 12 column for the dissociation constant measurement.

NEM modification of Hb did not alter the L band of Hb (Figure 7B). However, EAF PEGylation of Hb (Figure 7A) markedly decreased the L band of the proteins. Interestingly, reversible protection of Cys-93(β) restores the ellipticity of the L band of the EAF-PEG-Hb to the level of the unmodified Hb. Because the EAF-PEG-Hb can be taken as NEM conjugated with the PEG chains, it follows that the PEG chains conjugated at Cys93(β) unfavorably alter the interactions between the heme and the surrounding globin chain. The Soret band of Hb is informative of the interactions of the heme prosthetic group with the surrounding aromatic residues (29). It should be mentioned that all the Hb samples in the present study were in the oxy form and in the ferrous state. In the Soret region, the ellipticity with the maximal at 421 nm of HbA increased upon EAF PEGylation (Figure 7A) and NEM mediated modification (Figure 7B). However, the ellipticity of the EAF-PEG-Hb is slightly lower than that of the NEM modified Hb. In contrast, the ellipticity of the EAFPEG-Hb decreased upon reversible protection of Cys-93(β). Thus, the reversible protection of Cys-93(β) reduced the perturbation of the heme environment of the EAF-PEG-Hb. Intrinsic Front-Face Fluorescence of the PEG-Hb. Front-face fluorescence spectroscopy of Hb has been used to monitor the quaternary changes of Hb upon chemical modification of the protein. The fluorescence sensitivity has been assigned mostly to Trp-37(β) and reflects the conformational changes at the R1β2 interface of Hb. When excited at 280 nm (Figure 8A), the fluorescence intensity of HbA (curve a) decreased upon EAF PEGylation (curve b) and the NEM modification (curve d). In contrast, reversible protection of Cys-93(β) increases the fluorescence intensity of the PEG-Hb (curve c) and the NEM modified Hb (curve e). Thus, the quaternary changes of the EAF-PEG-Hb around the R1β2 interface is primarily a consequence of maleimide modification at Cys-93(β) and not a result of the PEG chains. O2 Affinity of the PEG-Hb. The P50 of HbA decreases from 14.4 to 6.9 mmHg upon EAF PEGylation, along with a decrease in the Hill coefficient (Table 2). The NEM modified thiolated Hb control showed a P50 (oxygen affinity) comparable to the EAF-PEG-Hb. This suggests that the PEG chains play a limited role in dictating the oxygen affinity of the EAF-PEG-Hb. However, the P50 of the PEG-Hb and the NEM modified Hb increased upon reversible protection of Cys-93(β), along with a slight decrease in the Hill coefficient. The PEG-Hb and the NEM modified Hb are essentially insensitive to the allosteric effector, 2,3-DPG. On the other hand, the adducts with reversible protection of Cys-93(β) exhibit a small but noticeable sensitivity to 2,3-DPG.

Figure 7. Circular dichroism spectra of the PEG-Hb (A), the NEM modified thiolated Hb (B), and the PEG-RR-Hb (C). Circular dichroic spectra of HbA (a), the PEG-Hb (b), the PEG-Hb with free Cys-93(β) (c), HbA (d), the NEM modified thiolated Hb (e), the NEM modified thiolated Hb with free Cys-93(β) (f), RR-Hb (g), the PEG-RR-Hb (h), and the PEG-RR-Hb with free Cys-93(β) (i). The Hb samples were recorded at 25 °C with a 0.2 cm light path cuvette (310 µL).

Autoxidation and Heme Loss of the PEG-Hbs. The autoxidation rate of PEG-Hb is an integrated result of the maleimide modification of the thiolated Hb and the conjugated PEG chains (16). Therefore, the NEM modified thiolated Hb was used as a control to investigate the autoxidation and the heme loss of the PEG-Hb. Autoxidation of the PEG-Hb in PBS Buffer. As shown in Table 3, the autoxidation rate constant (kox) of HbA increased upon NEM modification and further increased upon conjugation of the PEG chains. The kox of the hexaPEG-Hb is marginally higher than that of the diPEG-Hb site-specifically modified at Cys-93(β) (16). The kox for the PEG-Hb and the NEM modified Hb decreased significantly upon reversible protection of Cys-93(β). Accordingly, PEGylation of Cys-93(β) makes a significant contribution to increasing the autoxidation of the PEG-Hb.

Second Generation of EAF-PEG-Hbs

Bioconjugate Chem., Vol. 20, No. 11, 2009 2067

Figure 8. Intrinsic fluorescence emission spectra of the PEG-Hbs. (a) Hb; (b) the PEG-Hb without free Cys-93(β); (c) the PEG-Hb with free Cys93(β); (d) the NEM modified Hb without free Cys-93(β); (e) the NEM modified Hb with free Cys-93(β); (f) RR-Hb; (g) the PEG-RR-Hb without free Cys-93(β); (h) the PEG-RR-Hb with free Cys-93(β). The excitation wavelength for the fluorescence emission was 280 nm. The measurements were performed using a Shimadzu spectrofluorimeter at room temperature at Hb concentration of 6.0 mg/mL in PBS, pH 7.4. Table 2. Functional Properties of the PEG-Hbsa P50 (mmHg)/n sample

no effector

2.5 mM DPG

HbA EAF-PEG-Hb without free Cys-93(β) EAF-PEG-Hb with free Cys-93(β) EAF-NEM-Hb without free Cys-93(β) EAF-NEM-Hb with free Cys-93(β) RR-Hb EAF-PEG-RR-Hb without free Cys-93(β) EAF-PEG-RR-Hb with free Cys-93(β)

14.4/2.8 6.9/2.1 8.5/2.0 7.0/2.0 8.5/2.1 30.5/2.4 9.9/1.7 18.5/1.7

22.0/2.6 7.0/2.2 11.6/1.7 7.3/2.0 10.0/1.9 36.4/2.0 10.5/1.7 21.5/1.5

Figure 9. Kinetics of the heme transfer from the oxidized Hb samples to albumin. (A) PEG-Hb. HbA (a), the PEG-Hb (b), the NEM modified thiolated Hb (c), the PEG-Hb with free Cys-93(β) (d), and the NEM modified thiolated Hb with free Cys-93(β) (e). (B) PEG-RR-Hb. RRHb (f), the PEG-RR-Hb (g), and the PEG-RR-Hb with reversible protection of Cys-93(β) (h). Percent metheme-albumin (MHA) ) [MHA]/([metHb] + [MHA]) × 100%. Table 4. Fast Heme Loss Rate Constant of the PEG-Hbs

a Oxygen equilibrium curves of the samples were measured at 37 °C in PBS (pH 7.4) at Hb tetramer concentration of 0.5 mM. The Hill coefficient is given in the parentheses. P50 is the partial oxygen pressure at 50% saturation and is expressed in mmHg on the left of the Hill coefficient.

Table 3. Autoxidation of the PEG-Hb at 37 °Ca sample

kox (h-1)

HbA (SP-PEG5K)2-Hbb EAF-PEG-Hb without free Cys-93(β) EAF-PEG-Hb with free Cys-93(β) (SE)2-Hbb EAF-NEM-Hb without free Cys-93(β) EAF-NEM-Hb with free Cys-93(β) RR-Hb (SP-PEG5K)2-RR-Hbb EAF-PEG-RR-Hb without free Cys-93(β) EAF-PEG-RR-Hb with free Cys-93(β)

0.012 0.027 0.036 0.024 0.019 0.027 0.017 0.018 0.031 0.042 0.029

a

The initial pO2 of the Hb samples is around 150 mmHg before the autoxidation experiment and decreases during the process of the autoxidation. b Data obtained from Hu et al. (16).

Heme Loss of the PEG-Hb. Oxidation of ferrous heme iron to the ferric state weakens its fifth coordinate bond to proximal His residue and thereby increases the probability of the heme loss (30). The heme loss rate (Figure 9A) and the fast-phase heme loss rate

a

sample

kfasta (min-1)

HbA (SP-PEG5K)2-Hbb EAF-PEG-Hb without free Cys-93(β) EAF-PEG-Hb with free Cys-93(β) (SE)2-Hbb EAF-NEM-Hb without free Cys-93(β) EAF-NEM-Hb with free Cys-93(β) RR-Hb (SP-PEG5K)2-RR-Hbb EAF-PEG-RR-Hb without free Cys-93(β) EAF-PEG-RR-Hb with free Cys-93(β)

0.043 0.100 0.160 0.118 0.161 0.187 0.141 0.020 0.036 0.064 0.042

Fast heme loss rate constant. b Data obtained from Hu et al. (16).

constant (kfast) of Hb (Table 4) increased significantly upon EAF PEGylation. Reversible protection of Cys-93(β) significantly decreased the heme loss rate and the kfast of the PEG-Hb. Accordingly, PEGylation of Cys-93(β) significantly increased the heme loss of the PEG-Hb. Moreover, the NEM modified Hb showed a slightly higher heme loss rate and kfast than its corresponding PEG-Hb. Therefore, the conjugated PEG chains can increase the heme stability of the PEG-Hb. Autoxidation of the PEG-Hb in the Plasma. The PEG-Hb (40 mg/mL) is mixed with the plasma at equal volume to mimic the in ViVo 50% exchange transfusion. As seen earlier with PEGHb (16), kox of the PEG-Hb is attenuated in the plasma as compared to that in PBS buffer (Table 5). The reversible

2068 Bioconjugate Chem., Vol. 20, No. 11, 2009

Li et al.

Table 5. Autoxidation of the PEG-Hb in Plasma at 37 °C koxa (h-1) sample

in PBS bufferb

in plasmac

HbA EAF-PEG-Hb without free Cys-93(β) EAF-PEG-Hb with free Cys-93(β) RR-Hb EAF-PEG-RR-Hb without free Cys-93(β) EAF PEG-RR-Hb with free Cys-93(β)

0.019 0.047 0.025 0.029 0.070 0.042

0.009 0.023 0.016 0.015 0.040 0.025

a Autoxidation rate constant measured at 37 °C. b Hb concentration is at 20 mg/mL in PBS buffer, pH 7.4 c Hb concentration is at 20 mg/mL in the mixture of equal volumes of PBS buffer (pH 7.4) and bovine plasma.

protection of Cys-93(β) further reduces the kox of the EAFPEG-Hb in the plasma. Design of the Second-Generation EAF-PEG-Hb, the EAF-PEG-rr-Hb. Characterization of the EAF-PEG-RR-Hb with ReVersible Protection of Cys-93(β). The PEG-RR-Hb (curve h, Figure 2B) shows a higher hydrodynamic volume than RR-Hb (curve f). Besides, the PEG-RR-Hb shows a more symmetrical elution pattern than the PEG-Hb (curve c). The EAF PEGylated bis-thiopridyl RR-Hb (curve i) exhibits a slightly smaller hydrodynamic volume than the PEG-RR-Hb (curve h) and no essential change upon releasing the thiopyridyl groups from Cys93(β) (curve j). RP-HPLC patterns of the EAF-PEG-RR-Hb are presented in Figure 3B. The R-globin of Hb was right-shifted upon RRfumaryl cross-link (curve e). The EAF-PEG-RR-Hb (curve f) has essentially no unmodified β-globin and small amount of RR-globin, along with the presence of mono-, di-, and triPEGylated globins. For the PEGylated bisthiopyridyl RR-Hb, the unPEGylated β-globin increased with the peak left-shifted, whereas the RR-globin decreased (curve g). Upon releasing the thiopyridyl moiety from Cys-93(β), the peak corresponding to the unPEGylated β-globin is right-shifted (curve h). SDS-PAGE pattern of RR-Hb shows a new band with an apparent molecular weight of 32 kDa (lane 5, Figure 4). Some smaller additional bands are observed, possibly due to the presence of a small amount of impurity that is not well-separated from RR-Hb. It may be noted that the sample was not heated at 60 °C after chromatographic purification as done by Baxter to increase the purity of the RR-Hb. EAF PEGylation of RR-Hb generated several new bands (about six to eight) (lane 6). Reversible protection of Cys-93(β) did not essentially alter the pattern of the EAF-PEG-RR-Hb (lane 7), except for the intensity of the mono- and diPEGylated β-globin. The molecular radius of the EAF-PEG-RR-Hb is slightly higher than that of the EAF-PEG-Hb and further slightly decreased upon reversible protection of Cys-93(β) (Table 1). Structure of the EAF-PEG-RR-Hb. The reactivity of Cys93(β) of RR-Hb is lower than that of HbA (Figure 5B). The EAF-PEG-RR-Hb with reversible protection of Cys-93(β) carries two thiols per tetramer (curve f). EAF PEGylation targeted exclusively to the amino groups has little influence on the reactivity of Cys-93(β) of RR-Hb (curve e). The PEG-RR-Hbs generated by reductive alkylation (curve g) and acylation chemistry (curve h) based PEGylation exhibit higher reactivity of Cys-93(β) than RR-Hb and the EAF-PEG-RR-Hb. The presence of the RR-intramolecular cross-link slightly lowers the ellipticity of Hb in the Soret CD region (Figure 7C). EAF PEGylation of RR-Hb increases its ellipticity in the Soret region, reflecting the structural perturbation of the heme pocket upon PEGylation. However, reversible protection of Cys-93(β) can reduce this structural perturbation of the EAF-PEG-RRHb. The fluorescence studies (Figure 8B) further confirm that reversible protection of Cys-93(β) could stabilize the quaternary structure of the EAF-PEG-RR-Hb.

Autoxidation, Heme Loss, and Oxygen Affinity of the EAF-PEG-RR-Hb. Both RR-intramolecular cross-linking and EAF PEGylation significantly influence the kox of Hb (Table 3). When the two modifications are introduced into Hb simultaneously, the increase in kox is not additive. Presumably, the PEGylation induced increase in the kox of Hb is partly due to the destabilization of Hb tetramers. However, this destabilization is counteracted by the RR-intramolecular cross-link. On the other hand, the reversible protection of Cys-93(β) reduces the kox of the EAF-PEG-RR-Hb. The presence of RR-fumaryl cross-link in Hb and the PEG-Hb reduces the kfast by 53% and 60%, respectively (Table 4). Reversible protection of Cys-93(β) further increases the heme stability of the EAF-PEG-RR-Hb (Figure 9B and Table 4). The EAF-PEG-RR-Hb shows lower oxygen affinity than the EAF-PEG-Hb (Table 2). The oxygen affinity of the EAF-PEGRR-Hb exhibits little sensitivity to the presence of 2,3-DPG. In contrast, it shows some sensitivity to the presence of 2,3-DPG upon protection of Cys-93(β). Thus, the quaternary structural features of Hb are conserved better in the EAF-PEG-RR-Hb upon protection of Cys-93(β). The high O2 affinity of the EAFPEG-Hb comes from the site-specific PEGylation of Cys-93(β) and EAF PEGylation of the amino groups. Introduction of RRfumaryl cross-bridge and protection of Cys-93(β) could partially lower the O2 affinity of the EAF-PEG-Hb.

DISCUSSION EAF PEGylation of Hb was developed to simplify the surface decoration of Hb with the PEG chains and develop a nonvasoactive Hb derivative (9). This new PEGylation platform is a cost-effective protocol as exemplified in the development of MP4 by Sangart (31). Subsequent studies have established that the extension arm sandwiched between the amino groups and the PEG chains affords a unique structural advantage with Hb over the direct PEGylation platforms. EAF PEGylation has little impact on the tetramer stability of the PEG-Hb. In contrast, direct PEGylation destabilizes the quaternary structure of the PEG-Hb. We have now generated the hexaPEG-Hbs using the four PEGylation platforms, namely, EAF PEGylation (9), reductive alkylation (26), active ester chemistry (27), and thiocarbamoylation chemistry based PEGylation (32). The hexaPEG-Hbs generated by direct PEGylation exist essentially as Hb dimers (26, 27). On the other hand, the EAF-PEG-Hb is present predominantly as tetrameric species (Figure 5). The preclinical and clinical studies of MP4 have established that it does not induce any undesirable toxicity (31). Accordingly, the absence of toxicity in MP4 and the reported toxicity in Enzon PEG-bovine-Hb implicated by Baldwin et al. (33) may be related to the anticipated low tetramer stability of the Enzon product. Recently, Kluger and his associates (34) have shown that EAF PEGylation increased the nitrite reductase activity of Hb, predominantly due to PEGylation of Cys-93(β). They suggested that this activity is responsible for the vasoinactivity of the PEGHb. Accordingly, avoiding PEGylation of Cys-93(β) may result in the vasoactivity of PEG-Hb. However, Cys-93(β) is free in the vasoinactive PEG-bovine-Hb (Enzon product) that was prepared using acylation chemistry based PEGylation. Besides, the Euro PEG-Hb, a product prepared using EAF PEGylation under deoxy condition to avoid PEGylation of Cys-93(β), was also vasoinactive (35, 36). Thus, the vasoinactivity of the PEGHb does not correlate to the presence of PEGylation of Cys93(β). Avoiding PEGylation of Cys-93(β) has a limited influence on the oxygen affinity of the EAF-PEG-Hb. This further confirms that the high oxygen affinity of the EAF-PEG-Hb is an intrinsic consequence of PEGylation and not a direct result of PEGylation of Cys-93(β). The same conclusion has been

Second Generation of EAF-PEG-Hbs

Bioconjugate Chem., Vol. 20, No. 11, 2009 2069

Table 6. Site Selectivity of EAF PEGylation in Three Preparations modified residue

MP4a

Euro PEG-Hbb

(SP-PEG5K)6-Hbc

Cys-93∉(β) Lys-40 (R) Lys-120 (β) Lys-61 (R) Lys-7 (R) Lys-8 (β) Val-1 (β) Lys-11 (R) Lys-16 (R) Lys-56 (R) Lys-60 (R) Lys-90 (R) Lys-99(R) Lys-139 (R) Lys-17 (β) Lys-59 (β) Lys-61 (β) Lys-65 (β) Lys-66 (β) Lys-95 (β) Lys-132 (β) Lys-144 (β)

+ + + + + + + + + + -

+ + + + + + + + + + + + + + + +

+ + + + + + + + + + + +

a The data obtained from Vandegriff et al. (37). b The data obtained from Iafelice et al. (36). c The data obtained from Manjula et al. (9).

drawn previously based on EAF PEGylation of rHb[Cys93(β)fAla] (10). We have also carried out the EAF PEGylation of bis thiopyridyl Hb under deoxy condition, isolated the PEGylated product, and released the thiopyridyl moiety. This EAF-PEG-Hb prepared under deoxy conditions exhibited an O2 affinity comparable to the product generated under oxy condition. The site selectivity of (SP-PEG5K)6-Hb (9), MP4 (37), and Euro PEG-Hb (36) has been compared in Table 6. This confirms that EAF PEGylation of the amino groups is not quantitative except for PEGylation of Cys-93(β) in (SP-PEG5K)6-Hb and MP4. Even though Euro PEG-Hb was generated under deoxy conditions, the pattern of PEGylation is generally comparable to those of (SP-PEG5K)6-Hb and MP4. Moreover, Lys-40(R) is modified under the oxy condition, but not under the deoxy condition. Thus, PEGylation of Lys-40(R) may marginally influence the oxygen affinity of Hb. In addition, the amino groups of Hb accessible for EAF PEGylation are overlapping irrespective of the oxy or deoxy conditions used for PEGylation. Accordingly, we speculate that the hydrated PEG shell of the EAF-PEG-Hb with free Cys-93(β) leads to a preferential stabilization of the R state over the T state of Hb. This is the primary molecular aspect that increases the O2 affinity of Hb on PEGylation. The site-specific PEGylation of Cys-93(β) increases the O2 affinity of Hb (15), predominantly due to maleimide modification of Cys-93(β). These two molecular aspects are not additive and appear to significantly overlap each other. The high O2 affinity of MP4 is considered a desirable feature for HBOC, in terms of achieving a targeted O2 delivery to the hypoxic areas and reducing the vasoactivity with the autoregulatory mechanism. Our present data establish that the reversible protection of Cys-93(β) of un-cross-linked Hb does not significantly decrease the oxygen affinity of the PEG-Hb (Table 2). Accordingly, design of new versions of MP4 with free Cys93(β) essentially does not have an impact on the concepts of the new paradigms for the design of the HBOCs. This will provide an opportunity to evaluate the potential merits of avoiding PEGylation of Cys-93(β) in terms of the efficacy of the EAF-PEG-Hb. Besides, the new version of the EAF-PEGHbs with free thiols may also have an impact on counteracting the reperfusion injury during transfusion. The optimum level of O2 affinity of a PEG-Hb to achieve adequate tissue oxygenation without activating the autoregula-

tory principles has been the subject of intense study in recent years (17, 35, 38, 39). The Ajinomoto PEG-Hb (PHP) exhibits an O2 affinity lower than that of unmodified Hb. However, the vasoinactivity of PHP is not comparable to that seen with Enzon PEG-bovine-Hb (40). The PEG-Hb with an oxygen affinity close to that of unmodified Hb (P50 around 15 mmHg) is being considered as a better choice for achieving adequate tissue oxygenation. The O2 affinity of the PEG-RR-Hb with free Cys93(β) is higher than that of RR-Hb but slightly lower than that of HbA. EAF PEGylation of RR-Hb is less efficient in increasing the O2 affinity of RR-Hb. Presumably, the low oxygen affinity of RR-Hb is partially counteracted by the PEG shell to stabilize the oxy state of Hb. The RR-Hb has been PEGylated by us using three PEGylation platforms, reductive alkylation chemistry, acylation chemistry, and the present modified EAF PEGylation. Comparison of the reactivity of Cys-93(β) suggests that the oxy conformation of Hb is stabilized to a maximum level by the propyl-PEG chains of (Propyl-PEG5K)6-Hb, whereas to a minimal level by the EAF PEGylation. In the reductive alkylation based PEGylation, four of the six PEG chains are targeted to the four R-amino groups of Hb, while in the other two protocols, the PEG chains are targeted exclusively to the ε-amino groups. PEGylation on the R-amino groups is most effective in stabilizing the oxy conformation, while PEGylation of ε-amino by propionyl PEG chains reduces this propensity of the PEG shell. The oxygen affinity of the three PEG-RR-Hbs can reflect the stabilization of the oxy conformation by the PEG chains and correlates well with the reactivity of Cys-93(β). Engineering the extension arm between the ε-amino groups and the PEG chains further reduces this efficiency of the PEG shell to stabilize the R state of Hb. Many intramolecularly cross-linked and/or oligomerized Hbs have been designed as potential HBOC. However, their application as oxygen therapeutics has been impeded due to their vasoactivity and/or autoxidation mediated toxicities. PEGylation of Hb has emerged as a new chemical approach to overcome the vasoactivity of acellular Hb. EAF PEGylation is a new cost-effective approach to neutralize their vasoactivity. Reversible protection of Cys-93(β) during EAF PEGylation has many beneficial effects on the EAFPEG-Hbs. In the present study, EAF PEGylation of RR-Hb has been investigated as a second generation of the EAF-PEG-Hb with lower oxygen affinity. This modified EAF PEGylation approach is expected to integrate the structural and functional advantages of avoiding the PEGylation of Cys-93(β). On the other hand, it is expected to conserve the potential of PEGylation to neutralize the vasoactivity of Hb and have minimal perturbation of the quaternary structural aspects of Hb.

ACKNOWLEDGMENT This work was supported by a grant-in-aid from the American Heart Association Heritage Affiliate 9951021T, the National Institutes of Health grants HL58247 and HL71064, and the U.S. Army grant PR023085.

LITERATURE CITED (1) Harris, J. M., and Chess, R. B. (2003) Effect of PEGylation on pharmaceuticals. Nat. ReV. Drug DiscoVery 2, 214–221. (2) Pasut, G., Guiotto, A., and Veronese, F. M. (2004) Protein, peptide and non-peptide drug PEGylation for therapeutic application. Expert Opin. Ther. Pat. 14, 859–894. (3) Roberts, M. J., Bentley, M. D., and Harris, J. M. (2002) Chemistry for peptide and protein PEGylation. AdV. Drug DeliVery ReV. 54, 459–476. (4) Chang, T. M. (1999) Future prospects for artificial blood. Trends Biotechnol. 17, 61–67. (5) Klein, H. G. (2000) The prospects for red-cell substitutes. New Engl. J. Med. 342, 1666–1668.

2070 Bioconjugate Chem., Vol. 20, No. 11, 2009 (6) Conover, C. D., Linberg, R., Shum, K. L., and Shorr, R. G. (1999) The ability of polyethylene glycol conjugated bovine hemoglobin (PEG-Hb) to adequately deliver oxygen in both exchange transfusion and top-loaded rat models. Artif. Cells, Blood Substitutes, Immobilization Biotechnol. 27, 93–107. (7) Winslow, R. M., Gonzales, A., Gonzales, M. L., Magde, M. D., McCarthy, M., Rohlfs, R. J., and Vandegriff, K. D. (1998) Vascular resistance and efficacy of red cell substitutes in a rat hemorrhage model. J. Appl. Physiol. 85, 993–1003. (8) Acharya, A. S., Intaglietta, M., Tsai, A. G., Malavalli, A., Vandegriff, K., Winslow, R. M., Smith, P. K., Friedman, J. M., and Manjula, B. N. (2005) Enhanced molecular volume of conservatively PEGylated Hb: (SP-PEG5K)6-Hb is non-hypertensive. Artif. Cells, Blood Substitutes, Immobilization Biotechnol. 33, 239–255. (9) Manjula, B. N., Tsai, A. G., Intaglietta, M., Tsai, C. H., Ho, C., Smith, P. K., Perumalsamy, K., Kanika, N. D., Friedman, J. M., and Acharya, A. S. (2005) Conjugation of multiple copies of polyethylene glycol to hemoglobin facilitated through thiolation: Influence on hemoglobin structure and function. Protein J. 42, 133–146. (10) Li, D., Ho, N. T., Simplaceanu, V., Ho, C., Acharya, S. A., and Manjula, B. N. (2007) Molecular aspects of the high oxygen affinity of non-hypertensive hexaPEGylated Hb, (SP-PEG5K)6Hb. Artif. Cells, Blood Substitutes, Immobilization Biotechnol. 35, 107–118. (11) Li, D., Manjula, B. N., and Acharya, S. A. (2006) Extension arm facilitated PEGylation of proteins: correlation of the structure of PEGylated Hemoglobin with the extent of PEGylation. Protein J. 25, 263–274. (12) Muirhead, H., Cox, J. M., Mazzarella, L., and Perutz, M. F. (1967) Structure and function of haemoglobin. 3. A threedimensional fourier synthesis of human deoxyhaemoglobin at 5.5 Angstrom resolution. J. Mol. Biol. 28, 117–156. (13) Cheng, Y., Shen, T. J., Simplaceanu, V., and Ho, C. (2002) Ligand binding properties and structural studies of recombinant and chemically modified hemoglobins altered at beta 93 cysteine. Biochemistry 41, 11901–11913. (14) Juszczak, L. J., Manjula, B. N., Bonaventura, C., Acharya, S. A., and Friedman, J. M. (2002) UV resonance Raman study of β-93 modified hemoglobin A: chemical modifier-specific effects and added influence of attached poly(ethylene glycol) chains. Biochemistry 41, 376–385. (15) Manjula, B. N., Tsai, A., Upadhya, R., Perumalsamy, K., Smith, P. K., Malavalli, A., Vandegriff, K. D., Winslow, R. M., Intaglietta, M., Prabhakaran, M., Friedman, J. M., and Acharya, A. S. (2003) Site-specific PEGylation of hemoglobin at cys93(β): correlation between the colligative properties of the PEGylated protein and the length of the conjugated PEG chain. Bioconjugate Chem. 14, 464–472. (16) Hu, T., Li, D., Manjula, B. N., and Acharya, S. A. (2008) Autoxidation of the site-specifically PEGylated hemoglobins: role of the PEG chains and the sits of PEGylation in the autoxidation. Biochemistry 47, 10981–10990. (17) Acharya, S. A., Intaglietta, M., and Tsai, A. G. (2007) Modulation of the polyethylene glycol-hemoglobin structure to increase the efficiency of plasma expansion and O2 carrying capacity. Trans. Altern. Trans. Med. 9, 254–264. (18) Manjula, B. N., and Acharya, A. S. (2003) Purification and molecular analysis of hemoglobin by high-performance liquid chromatography. Methods Mol. Med. 82, 31–47. (19) Chatterjee, R., Welty, E. V., Walder, R. Y., Pruitt, S. L., Rogers, P. L., Arnone, A., and Walder, J. A. (1986) Isolation and characterization of a new hemoglobin derivative cross-linked between the R chains (lysine 99R1-lysine 99R2). J. Biol. Chem. 261, 9929–9937. (20) Benesch, R. E., and Kwong, S. (1990) The stability of the heme-globin linkage in some normal, mutant, and chemically modified hemoglobins. J. Biol. Chem. 265, 14881–14885. (21) Benesch, R. E. (1994) The stability of the heme-globin linkage: measurement of heme exchange. Methods Enzymol. 231, 496– 502.

Li et al. (22) Di Iorio, E. E. (1981) Preparation of derivatives of ferrous and ferric hemoglobin. Methods Enzymol. 75, 57–72. (23) Manning, L. R., Jenkins, W. T., Hess, J. R., Vandegriff, K., Winslow, T. M., and Manning, J. M. (1996) Subunit dissociations in natural and recombinant hemoglobins. Protein Sci. 5, 775–781. (24) Hu, T., Prabhakaran, M., Acharya, S. A., and Manjula, B. N. (2005) Influence of the chemistry of conjugation of poly(ethylene glycol) to Hb on the oxygen-binding and solution properties of the PEG-Hb conjugate. Biochem. J. 392, 555–564. (25) Ampulski, R. S., Ayers, V. E., and Morell, S. A. (1969) Determination of the reactive sulfhydryl groups in heme proteins with 4,4′-dipyridine disulfide. Anal. Biochem. 32, 136–169. (26) Hu, T., Manjula, B. N., Li, D., Brenowitz, M., and Acharya, A. S. (2007) Influence of intramolecular cross-links on the molecular, structural and functional properties of PEGylated hemoglobin. Biochem. J. 402, 143–151. (27) Li, D., Hu, T., Manjula, B. N., and Acharya, S. A. (2008) Non-conservative surface decoration of hemoglobin. Biochim. Biophys. Acta 1784, 1395–1401. (28) Zentz, C., Pin, S., and Alpert, B. (1994) Stationary and timeresolved circular dichroism of hemoglobins. Methods Enzymol. 232, 247–266. (29) Hsu, M. C., and Woody, R. W. (1971) The origin of the heme Cotton effects in myoglobin and hemoglobin. J. Am. Chem. Soc. 93, 3515–3525. (30) Bunn, H. F., and Jandl, J. H. (1968) The renal handling of hemoglobin. J. Biol. Chem. 243, 465–475. (31) Young, M. A., Malavalli, A., Winslow, N., Vandegriff, K. D., and Winslow, R. M. (2007) Toxicity and hemodynamic effects after single dose administration of malPEG-hemoglobin (MP4) in rhesus monkeys. Transl. Res. 149, 333–342. (32) Acharya, S. A, Smith, P. K., and Manjula, B. N. (2006) PEGylated non-hypertensive hemoglobins, methods of preparing same and used there off, U. S. Patent 7,144,989. (33) Baldwin, A. L., Wilson, L. M., and Valeski, J. E. (1998) Ultrastructural effects of intravascularly injected polyethylene glycol-hemoglobin in intes-tinal mucosa. Am. J. Physiol. 275, H615–H625. (34) Lui, F. E., Dong, P., and Kluger, R. (2008) Polyethylene glycol conjugation enhances the nitrite reductase activity of native and cross-linked hemoglobin. Biochemistry 47, 10773–10780. (35) Portoro, I., Kocsis, L., Herman, P., Caccia, D., Perrella, M., Ronda, L., Bruno, S., Bettati, S., Micalella, C., Mozzarelli, A., Varga, A., Vas, M., Lowe, K. C., and Eke, A. (2008) Towards a novel haemoglobin-based oxygen carrier: Euro-PEG-Hb, physico-chemical properties, vasoactivity and renal filtration. Biochim. Biophys. Acta 1784, 1402–1409. (36) Iafelice, R., Cristoni, S., Caccia, D., Russo, R., Rossi-Bernardi, L., Lowe, K. C., and Perrella, M. (2007) Identification of the sites of deoxyhaemoglobin PEGylation. Biochem. J. 403, 189–196. (37) Vandegriff, K. D., Malavalli, A., Mkrtchyan, G. M., Spann, S. N., Baker, D. A., and Winslow, R. M. (2008) Sites of modification of Hemospan, a poly(ethylene glycol)-modified human hemoglobin for use as an oxygen therapeutic. Bioconjugate Chem. 19, 2163–2170. (38) Cabrales, P., and Friedman, J. M. (2007) PEGylated hemoglobins: mechanisms to avoid vasoconstriction and maintain perfusion. Trans. Altern. Trans. Med. 9, 281–293. (39) Acharya, S. A., Acharya, V. N., Kanika, N. D., Tsai, A. G., Intaglietta, M., and Manjula, B. N. (2007) Non-hypertensive tetraPEGylated canine heamoglobin: correlation between PEGylation, O2 affinity and tissue oxygenation. Biochem. J. 405, 503–511. (40) Rohlfs, R. J., Bruner, E., Chiu, A., Gonzales, A., Gonzales, M. L., Magde, D., Jr., Vandegriff, K. D., and Winslow, R. M. (1998) Arterial blood pressure responses to cell-free hemoglobin solutions and the reaction with nitric oxide. J. Biol. Chem. 273, 12128–12134. BC900170E