phosphinic Acid - American Chemical Society

Jul 27, 2010 - Ramachandra S. Hosmane*,†,§. Laboratory for Drug Design and Synthesis, Department of Chemistry and Biochemistry, University of Maryl...
0 downloads 0 Views 3MB Size
1494

Bioconjugate Chem. 2010, 21, 1494–1507

Bis[2-(3-carboxyphenoxy)carbonylethyl]phosphinic Acid (m-BCCEP): A Novel Affinity Cross-Linking Reagent for the β-Cleft Modification of Human Hemoglobin Hongyi Cai,† Timothy A. Roach,† Margaret Dabek,† Karla S. Somerville,† Seetharama Acharya,‡ and Ramachandra S. Hosmane*,†,§ Laboratory for Drug Design and Synthesis, Department of Chemistry and Biochemistry, University of Maryland, Baltimore County, 1000 Hilltop Circle, Baltimore, Maryland 21250, and Albert Einstein College of Medicine, Jack and Pearl Resnick Campus, 1300 Morris Park Avenue, Bronx, New York 10461. Received March 3, 2010; Revised Manuscript Received June 18, 2010

The design and synthesis of bis[2-(3-carboxyphenoxy)carbonylethyl]phosphinic acid (m-BCCEP, 1) as a sitedirected affinity reagent for cross-linking human hemoglobin have been reported as part of our long-term goal to generate artificial blood for emergency transfusions. Molecular modeling techniques were used to design the reagent, employing crystal coordinates of human hemoglobin A0 imported from the Protein Data Bank. It was synthesized in four steps commencing from 3-hydroxybenzoic acid. The reagent 1 was converted to its trisodium salt to allow effective cross-linking in an aqueous medium. The reagent 1, as its trisodium salt, was found to specifically cross-link stroma-free human hemoglobin A0 in the β-cleft under oxygenated reaction conditions at neutral pH. The SDS-PAGE analyses of the modified hemoglobin pointed to the molecular mass range of 32 kDa as anticipated. The HPLC analyses of the product suggested that the cross-link had formed between the β1-β2 subunits. Molecular dynamics simulation studies on the reagent-HbA0 complex suggested that the predominant amino acid residues involved in the cross-linking are N-terminus Val-1 or Lys-82 on one of the β-subunits and Lys-144 on the other. These predictions were borne out by MALDI-TOF MS analyses data of the peptide fragments obtained from tryptic digestion of the cross-linked product. The data also suggested the presence of a minor cross-link between Val-1 and Lys-82 on the opposing subunits. The oxygen equilibrium measurements of the m-BCCEP-modified hemoglobin product at 37 °C showed oxygen affinity (P50 ) 25.8 Torr) comparable to that of the natural whole blood (P50 ) 27.0 Torr) and significantly lower than that of stroma-free hemoglobin (P50 ) 14.19 Torr) assayed under identical conditions. The measured Hill coefficient value of 1.91 of the m-BCCEPmodified Hb product points to the reasonable retainment of oxygen-binding cooperativity after the cross-link formation.

INTRODUCTION Current world events replete with war and terrorism necessitate countless emergency transfusions, resulting in scarcity of blood, especially of rare types (1). Even otherwise, limitations on storage stability of whole blood, the necessity for bloodtyping and cross-matching before transfusion, and the fear of possible transmission of infectious diseases upon transfusion, not to mention other psychological or religious reasons, reinforce the need to develop a safe and effective blood substitute (2-5). The latter must be universally acceptable without the requirement for typing or cross-matching, be free of pathogens, be storable for prolonged periods of time, and should allow transfusion at the scene of the accident without having to transport the patient to the nearest hospital (6-10). The efforts to replace human blood can be traced back to the early 1900s (11, 12). Despite extensive investigations in the past few decades, it still remains a daunting task to discover an optimal blood substitute for emergency transfusion (7, 13-16). While many approaches have been taken in the past to develop blood substitutes, including plasma expanders (17), perfluorocarbon (PFC) suspensions (18-20), and microencapsulation (21-24), the one based * To whom the correspondence should be addressed. Tel: 410-4553717. Fax: 410-455-1148. E-mail: [email protected]. † University of Maryland. ‡ Albert Einstein College of Medicine. § Recently retired.

on cell-free hemoglobins for an oxygen-carrying resuscitation fluid has excellent prospects since (a) hemoglobin solutions are completely metabolizable and are well-tolerated by the body and (b) hemoglobin is fully saturated with oxygen under ambient conditions, has oncotic activity, and exhibits cooperative oxygen binding characteristics (7, 14, 25-30). Hemoglobin is the ironcontaining oxygen-transport protein in the red blood cells (RBCs) of vertebrates (31). Nonetheless, approaches employing cell-free hemoglobin are plagued by two major problems, short retention time of hemoglobin in circulation (1-4 h) and high oxygen affinity that causes inadequate oxygen release into tissues. It is highly toxic to the kidney and causes hypertension (32). Removal of the RBC membrane stroma resulted in stromafree hemoglobin (SFHb) that had less renal toxicity in animals (33). The phase I clinical trials of SFHb still exhibited renal toxicity, as well as high blood pressure (34). Therefore, the use of native Hb was discontinued, and instead, modifications that allow Hb or SFHb function outside the RBCs have been extensively studied (15). When outside the RBCs, Hb tetramer breaks into two Rβ dimers due to the loss of its allosteric effector, 2,3-diphosphoglycerate (DPG) (35, 36), which electrostatically holds the two dimers together (see Figure 1). Since Rβ dimer is considerably smaller than the tetramer; it will be easily filtered out of the bloodstream and is consequently cleared by the kidney causing kidney toxicity (7, 37, 38). DPG is equipped with multiple negative charges at physiological pH and thus can bind to the positively charged amino acid residues residing on the β-sub-

10.1021/bc100113y  2010 American Chemical Society Published on Web 07/27/2010

Hemoglobin Cross-Linking Reagent for β-Cleft Modification

Bioconjugate Chem., Vol. 21, No. 8, 2010 1495

period. Considering that a number of otherwise promising reagents were rejected by industry because of perceived synthetic difficulty, poor yields, lack of prolonged stability, or exorbitant costs of production (2, 5), the observed properties of reagent m-BCCEP are indeed promising. We have demonstrated here that its trisodium salt cross-links human hemoglobin under aqueous, physiological conditions. Reverse-phase HPLC in conjunction with SDS-PAGE analyses suggested that the reagent is specific for cross-linking the β chains of hemoglobin. Figure 1. (A) Schematic representation of the tetrameric structure of hemoglobin formed by electrostatic interaction between the anionic charges of 2,3-diphosphoglycerate (2,3-DPG) and the cationic charges of amino acid residues present on the β-subunit of hemoglobin and (B) the molecular structure of 2,3-diphosphoglycerate.

units, forming a stable Hb tetramer with a central cavity called the DPG pocket or β-cleft (39). The absence of DPG also leads to high oxygen affinities of Hb, which prevents Hb from unloading sufficient amount of oxygen to tissues (40-42). Therefore, the current efforts to develop blood substitutes based on cell-free hemoglobin are directed toward (a) tuning its oxygen affinity to afford adequate oxygen delivery from lung to tissues via covalent cross-linking, preferably within the β-cleft, with an appropriate reagent that mimics the hemoglobin’s natural allosteric modifier 2,3-diphosphoglycerate (DPG) (43), and (b) increasing the steric bulk of the cross-linked hemoglobin through polymerization to allow its retention in circulation for prolonged periods of time (44, 45). The increased size would also prevent its facile sieving through the endothelial lining of blood vessels, where they scavenge nitric oxide (NO), the endothelium-derived relaxation factor (46-50). The latter is essential for relaxing the muscular walls surrounding the blood vessels, and therefore, its depletion results in elevated arterial pressure (46-50). Many laboratories, including ours, have designed and synthesized many intramolecular (interdimer) or intermolecular (intertetramer) cross-linking reagents containing bifunctional or polyfunctional moieties to chemically modify cell-free Hb (5, 13, 51-75). Different degrees of success have been achieved, and a few products have even gone on to clinical trials (4, 7). However, these modified hemoglobins still suffer from one or the other problems, including too high oxygen affinity, too low intravascular retention time, too facile autoxidation of Fe2+ to Fe3+ in Hb, forming metHb that lacks the capacity to carry oxygen, or other incompatible physiological characteristics (15, 76, 77). The challenge remains to synthesize an ideal reagent which closely mimics DPG to the extent that it would be specifically attracted to the β-cleft and cause the least conformational change in the native structure of Hb upon subsequent cross-linking and polymerization. We report here the design, synthesis, and functional studies of bis[2-(3-carboxyphenoxy)-carbonylethyl]phosphinic acid (m-BCCEP, 1), a novel affinity cross-linking reagent for the

β-cleft modification of human hemoglobin. Synthesis of this reagent was accomplished in four steps starting from commercially available, inexpensive materials and can easily be scaled up. The reagent as well as its trisodium salt are crystalline solids that are stable at ambient temperature for an indefinite

EXPERIMENTAL PROCEDURES Materials and Methods. IR spectra were obtained on a Thermo Nicolet Avatar 370 FT-IR spectrometer. The NMR spectra were obtained on an Oxford AS 400 (400 MHz) spectrometer. The mass spectra were acquired on either an Esquire 3000 plus ESI mass spectrometer (Bruker Daltonics) or an Autoflex MALDI TOF/TOF MS instrument (Bruker Daltonics).The purity of samples of newly synthesized materials was assessed by a combination of 400 MHz NMR spectroscopy, mass spectrometry, thin layer chromatography, and elemental microanalyses. The latter data were obtained from Atlantic Microlab, Inc., Norcross, Georgia. Molecular Modeling Studies. Molecular modeling was performed on a Silicon Graphics workstation, using the software Insight/Discover (Accelrys, San Diego, CA, USA). The X-ray coordinates of human Hb (78), imported from the Brookhaven National Laboratory, Upton, NY, USA, were employed for molecular modeling studies. It was energy-minimized and docked into the β-cleft (DPG pocket) of hemoglobin, followed by energy minimization of the reagent-Hb complex. All atoms that were 12 Å or farther from the ligand were fixed with a temperature constant of 300 K. No constraints were applied to the remaining residues in and around the ligand site. The complex was minimized to convergence using consecutive steepest descent and conjugate gradient (VAO9A) and NewtonRaphson energy minimization algorithms, with a final total Insight energy of 505.68 kals/mol, average absolute derivative of 0.000 129, standard deviation of absolute derivative of 0.000 114, and rms derivative of 0.000 172. In order to simulate the natural environment even further, a femtosecond molecular dynamics simulation (5000 iterations) at 300 K was performed on the above energy minimized protein-ligand complex by soaking the latter in an aqueous layer of 5 Å thickness all around the complex. There were no morse or cross terms. The distance ranges of the ester carbonyl carbon atoms of the reagent from the appropriate ε-amino nitrogen atoms of lysine residues or the R-amino nitrogen atom of the N-terminus valine were computed from graphs derived from the dynamics trajectories. Organic Synthesis. Commercial reagents were employed without further purification. Organic reagents and solvents were purchased from Fischer Scientific Co. or Aldrich Chemical Co., and solvents were dried before use. t-Butyl 3-Hydroxybenzoate (2). In a 250 mL round-bottom flask equipped with a magnetic stirrer, 3-hydroxybenzoic acid (4.6 g, 33.3 mmol) was dissolved in anhydrous benzene (200 mL), and the solution was heated to reflux. N,N′-Dimethylformamide di-t-butylacetal (13.54 g, 66.6 mmol) was added dropwise using a dropping funnel. The reaction mixture was allowed to reflux for 90 min or more. The solution was cooled down to room temperature and then washed with saturated aqueous NaHCO3 solution (2 × 300 mL). The organic layer was dried over anhydrous Na2SO4. Filtration followed by solvent removal under reduced pressure yielded 4.72 g (73%) of pure 2 as a colorless oil. 1H NMR (DMSO-d6): δ 9.6 (s, 1H, OH), 7.2-6.7 (m, 4H, Ar-H), 1.5 (s, 9H, t-Bu). Anal. Calcd for C11H14O3 · 0.2H2O: C, 66.78; H, 7.34. Found: C, 66.80; H, 7.32.

1496 Bioconjugate Chem., Vol. 21, No. 8, 2010

3-t-Butoxycarbonylphenyl Acrylate (3). To a solution of t-butyl 3-hydroxybenzoate (2) (4.23 g, 21.8 mmol) in anhydrous THF (120 mL) was added potassium t-butoxide (2.53 g, 22.5 mmol). The mixture was allowed to stir for 30 min at 0 °C in an ice bath, after which acryloyl chloride was added (2.15 mL, 26.5 mmol). The reaction mixture was allowed to stir overnight. After filtering the mixture, the THF was removed by rotary evaporation. Then, the mixture was dissolved in ethyl ether (100 mL) and washed with aq. HCl (0.1 N, 3 × 100 mL), followed by distilled water (3 × 100 mL). The organic layer was then dried with Na2SO4, filtered, and the filtrate evaporated to give a brownish-yellow oil (4.43 g). This oil was purified by silica gel chromatography, eluting with chloroform to give a yellow oil (3) (3.69 g, 68%). 1H NMR (CDCl3): δ 7.9-6.1 (m, 4H,

Ar-H), 6.7-6.6 (d, 1H, CHdCH2), 6.4-6.2 (dd, 1H, CHdCH2), 6.1-6.0 (d, 1H, CHdCH2), 1.6 (s, 9H, t-Bu). Anal. Calcd for C14H16O4: C, 67.73; H, 6.50. Found: 67.89; H, 6.51. Bis[2-(3-t-butoxycarbonyl)phenoxycarbonylethyl]phosphinic Acid (4). To an ice-cold, stirred solution of ammonium phosphinate (0.44 g, 5.25 mmol) in 50 mL dry dichloromethane was added N,O-bistrimethylsilylacetamide (3.74 g, 18.4 mmol) and the above acryloyl ester (3) (2.61 g, 10.5 mmol) while under argon atmoshere. The reaction mixture was allowed to stir overnight. It was partitioned between CH2Cl2 and H2O, and the organic layer was washed with aq. HCl (0.1 N, 3 × 100 mL). The organic layer was then dried over anhydrous Na2SO4, filtered, and the filtrate evaporated to give a brown oil. This oil was purified by silica gel chromatography, eluting with chloroform-methanol mixture (7.5:2.5) to give a brownishyellow oil (4) (2.30 g, 78%). 1H NMR (CDCl3): δ 7.90-7.77 (dd, 2H, Ar-H), 7.73-7.62 (m, 2H, Ar-H), 7.48-7.34 (m, 2H, Ar-H), 7.30-7.18 (d, 2H, Ar-H), 3.09-2.82 (m, 4H, CH2-PO), 2.38-2.09 (m, 4H, CH2-PO), 1.57 (s, 18H, t-Bu). Anal. Calcd for C28H35PO10 0.5H2O: C, 58.84; H, 6.34. Found: C, 58.44; H, 6.54. Bis[2-(3-carboxyphenoxy)carbonylethyl]phosphinic Acid (1). A solution of (4) (563 mg, 1 mmol) in 50 mL of dry dichloromethane was refluxed with 4 mL of trifluoroacetic acid for 3 h. After evaporation of the solvent on a rotavapor, a brownishyellow residue was obtained, which was washed with diethyl ether to yield 419 mg (93%) white solid compound (1). 1H NMR (DMSO-d6): δ 7.83-7.38 (m, 8H, Ar-H), 2.81-2.63 (m, 4H, CH2), 2.08-1.85 (m, 4H, CH2). 13C NMR (DMSO-d6): δ 174.10 (ester CdO), 167.85 (acid CdO), 157.92 (OsCdC), 132.54 (Ar-C), 130.11 (Ar-C),126.89 (Ar-C), 123.25 (Ar-C), 120.50 (Ar-C), 27.22 (CH2CdO), 25.07 (CH2PdO); IR (νmax) 3027-2650 (br, CO2H), 1753 (CdO), 1701 (CdO) cm-1; MS (ESI) m/z 451 (MH+). Anal. Calcd for C20H19PO10: C, 51.29; H, 4.52. Found: C, 51.26; H, 4.48.

Cai et al.

Conversion of Reagent 1 into the Corresponding Trisodium Salt. A glass column (30 cm ×100 cm) was packed with the ion-exchange resin AG 50W-X8 (H+ form) (Bio-Rad). The resin was converted to its sodium form by equilibration with 1 N NaOH until the PH was basic (pH 14 or more). The column was then thoroughly washed with deionized water until the pH of the eluent was neutral (or equal to the pH of water). Reagent 1 (135 mg, 0.3 mmol) was then dissolved into ethanol (10 mL) and loaded onto the top of the resin, and the column was eluted with 500 mL of deionized water. The combined fractions were frozen using liquid nitrogen and lyophilized for 2 days to yield 118 mg (77%) of a yellowish salt. This salt was used for hemoglobin modification without further purification. 1 H NMR (D2O): δ 7.81-7.26 (m, 8H, Ar-H), 2.95-2.87 (m, 4H, CH2), 2.11-1.97 (m, 4H, CH2). Reaction of m-BCCEP with Hemoglobin. Highly purified stroma-free hemoglobin A (SFHbA) in the oxygenated state was obtained from Walter Reed Army Institute of Research, Washington, DC. A stock solution (2 mM) of human hemoglobin was made by dilution of 2.8 mL concentrated hemoglobin (13.7 g/dL) in the oxygenated state at room temperature with 0.12 mL of 0.1 M phosphate buffer, pH 7.4. Next, into three 2 mL Eppendorf tubes was added 500 µL of the stock (2 mM) Hb solution and 1.16 mL of 0.1 M phosphate buffer, pH 7.4. A 150 mM stock reagent solution of m-BCCEP was prepared by dissolving 77.5 mg of m-BCCEP (trisodium salt form) into 1 mL of 0.1 M phosphate buffer, pH 7.4, and used immediately in the cross-linking reactions by adding 333 µL to two of the above Eppendorf tubes for a total volume of 2 mL per tube. The final concentration was 25 mM (50-fold excess) for the reagent and 0.5 mM for Hb. The third tube was used for control purposes and was brought to a volume of 2 mL with additional buffer. The reactions were then kept at temperature of 30 °C for 12 h and monitored by reverse-phase HPLC using a C4 column (conditions listed below) every 4 h. After that time, no further modifications were present in the chromatograms, and the solutions were frozen and stored at -81 °C until further analyses could be performed. HPLC Analysis and Purification of Modified Hemoglobin. The modified and native hemoglobin were analyzed by reversephase HPLC using a C4 Vydac 214TP1010 column (10 × 250 mm) with developer A (CH3CN, 0.1% TFA) and developer B (H2O, 0.1%TFA). The column was equilibrated with 35% developer A and 65% developer B and a gradient employed to give a final composition of 50% developer A and 50% developer B over a period of 90 min at 3 mL/min while monitoring at λmax 214 nm. The cross-linking solutions were passed through a Sephadex G-75 column (2.5 × 45 cm) equilibrated with 0.05 M Tris buffer, pH 7.4, 1 M MgCl2, and eluted with the same buffer to denature any un-cross-linked Hb. The fractions were combined and then simultaneously desalted, and the buffer was exchanged (0.05 M Tris, pH 7.4) using Millipore Biomax 50 concentrators. The concentrated Hb was then frozen at -81 °C until further use. SDS-PAGE Analysis of Modified Hemoglobin. The extent of cross-linking of the hemoglobin was determined by polyacrlamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS-PAGE). Modified hemoglobin (20 µL) and control hemoglobin (20 µL) were added into 40 µL LB Buffer separately (Laemmli sample buffer/2-mercaptoethanol ) 95:5) and heated to boil for 5 min, then loaded 10 µL per well and the gel run (12% polyacrylamide, 0.375 M Tris-HCl, pH 8.8) at 150 V for 35 min. The protein bands on the gel were then stained for 1 h with Coomassie Brilliant Blue dye and then destained using Bio-Rad destain. SDS-PAGE standards were used to compare the molecular weight.

Hemoglobin Cross-Linking Reagent for β-Cleft Modification

Figure 2. Energy minimized structure of reagent m-BCCEP.

Tryptic Digestion Analyses. The native and modified globin chains were purified for tryptic digestion by preparative reversephase HPLC using a C4 Vydac 214TP1022 column (22 × 250 mm). The column was equilibrated with 35% developer A (CH3CN, 0.1% TFA) and 65% developer B (H2O, 0.1% TFA) and a linear gradient employed to give a final composition of 50% developer A and 50% developer B over a period of 95 min at 13 mL/min while monitoring at λmax 214 nm. The appropriate fractions were then pooled and frozen in a liquid nitrogen bath followed by lyophilization to a powder. The isolated β (1.2 mg) and XL-β chains (1.2 mg) were then dissolved separately into 200 µL buffer (50 mL H2O, 0.1% SDS, 50 mM Tris-HCl, 5 µM β-mercaptoethanol at pH 8), and 60 µL of TPCK-trypsin (1 µg/ µL) was added (protein/trypsin ) 20:1 w/w). Digestion was carried out for 24 h at 37 °C with gentle agitation. Digestion was terminated by freezing the samples and then lyophilizing them to a powder. The digested globin chains were analyzed by reverse-phase HPLC using a

Bioconjugate Chem., Vol. 21, No. 8, 2010 1497

C18 Phenomenex column (4.5 × 250 mm, 5 µm, 300 A); the injection was 10 µL, and the column was equilibrated with 5% developer A (CH3CN, 0.1% TFA) and 95% developer B (H2O, 0.1%TFA) and a linear gradient employed to give a final composition of 50% developer A and 50% developer B over a period of 120 min at 1 mL/min while monitoring at λmax 214 nm. The fragments obtained were collected from the HPLC column, as they eluted out and lyophilized individually. The fractionated tryptic peptides were then analyzed by matrixassisted laser desorption/ionization time-of-flight MS (MALDITOF). Mass Spectral Analyses. The fractionated tryptic peptides were analyzed by MALDI-TOF-MS (79-81), using an Autoflex MALDI-TOF/TOF MS instrument. All experiments were performed using R-cyano-4-hydroxycinnamic acid as the matrix. Matrix (10 mg) was dissolved into 0.5 mL 7:3 H2O (0.1% TFA): CH3CN to make a final concentration of matrix as 20 mg/mL. Each lyophilized tryptic fragment was dissolved in 20 µL H2O, mixed well, and applied 0.5 µL onto a sample well, followed by an equal volume of the above matrix solution. One well is reserved for the matrix alone as a reference. The samples were allowed to dry at room temperature before placing into the spectrometer. Insulin (MW: 5734.51), ACTH (MW: 2465.1989), Angiotensin (MW: 1046.5423) and Insulin Oxidized B chain (MW: 3494.6513) were used to calibrate the machine. Oxygen Equilibrium (P50) Studies. Oxygen equilibrium curve (OEC) was obtained in the Hemox buffer (pH 7.4) at 37 °C using a Hemox-Analyzer (TCS Scientific Corp., Huntington Valley, PA). The sample was oxygenated with an increasing partial pressure of oxygen and deoxygenated with nitrogen gas. The analyzer detects the oxygen pressure with a Clark oxygen electrode and simultaneously calculates the oxygen saturation of Hb via a dual-wavelength spectropotometry. value for P50 (the oxygen pressure where Hb is half-saturated with oxygen) was obtained from the OEC. The Hill coefficient, which represents

Figure 3. Stereoview of the energy minimized complex of reagent m-BCCEP with HbA0.

Figure 4. Close-up stereoview of the energy-minimized m-BCCEP-HbA0 complex. Only those amino acid residues of SFHb that lie within a 12 Å radius from the reagent in the β-cleft are shown.

1498 Bioconjugate Chem., Vol. 21, No. 8, 2010

Cai et al.

Figure 5. Close-up monoview of the energy-minimized m-BCCEP-HbA0 complex shown in Figure 4, showing the amino acid residues (lysines or R-terminus valine) whose appropriate (ε- or R-) amino groups fall within a ∼9 Å distance from the two cross-linking sites of the reagent.

the cooperative effect of the four subunits in Hb molecule, was obtained from the slope of Hill plot (log(Y/(1 - Y) versus log P), where Y is fractional saturation of the Hb with oxygen and P is the oxygen pressure in Torr.

RESULTS AND DISCUSSION We have recently reported the synthesis and hemoglobin cross-linking studies of a reagent called BPPCEP (2) (Bis[2(4-phosphonooxyphenoxy)carbonylethyl]phosphinic acid) (59). The latter contains the same number (five) of anionic charges and the same number (two) of phosphate groups as DPG, but the reagent was so highly reactive that it produced an intractable mixture of products upon hemoglobin cross-linking. Three major products were laboriously isolated from the reaction mixture and fully characterized (59). While all three of them were found to be β-β cross-linked, their oxygen affinities, ranging from 12.5 to 15.5 Torr, were not superior to the oxygen affinity of the stroma-free hemoglobin that was used for cross-linking, which measured 13.1 Torr under identical conditions (59). Even worse was their oxygen-binding cooperativity characteristics as measured by the Hill coefficients, which ranged from 0.5 to 0.9 as compared to ∼2.5 for HbA0 (2, 5). Interestingly, we had discovered a decade earlier that a carboxylate group in place of a phosphate group as in reagent 3 (bis[2-(4-carboxyphenoxy)carbonylethyl]phosphinic acid (p-BCCEP)), despite possessing only three negative charges under physiological conditions, underwent a much cleaner reaction with hemoglobin, but unfortunately, the yield of the cross-linked product was only 15-20%, thus making it practically unsuitable or unattractive for further development (57). Our molecular modeling studies (see below) suggested that the movement of this carboxylate group from the para position to the meta would considerably enhance the interactions of the ligand with the protein, making worthwhile the synthesis and Hb cross-linking studies of the title reagent m-BCCEP (1). As revealed by the data presented herein, this indeed proved to be the case. Molecular Modeling Studies. Molecular modeling studies were performed on a Silicon Graphics workstation, using the X-ray coordinates of human HbA0 (78) imported from the Brookhaven National Laboratory, Upton, New York. Although

molecular modeling techniques have been used for extensive investigations of allosteric modifications of hemoglobin (82-85), drug delivery of hemoglobin bioconjugates (86) and biophysical characterization of genetic variants (87), mutants (88), selfassociation (89), and packing (90) of hemoglobins, the documented use of molecular modeling for the de novo design of hemoglobin cross-linking reagents is somewhat rare, other than our own earlier reports on the subject (57, 59, 74). The initial structural framework for the cross-linking reagent was laid down by assessment of the nature and dimension of the DPG pocket and the identification of the target amino acids on the diagonally opposed β1 and β2 subunits for cross-linking, as well as the tether length, geometry, and the electrophilic functional groups required for efficient cross-linking. To this end, the modeling studies were first carried out using DPG itself as the ligand. The reagent m-BCCEP (1) was built and energy-minimized employing the molecular modeling software Insight/DiscoVer (Accelrys Software, Inc., San Diego, California). The structure of the energy-minimized m-BCCEP is shown in Figure 2. The reagent is equipped with two activated ester functionalities to serve as the two cross-linking sites. The two end carboxylate groups and the central phosphinic hydroxyl are anticipated to provide the necessary anionic charges at biological pH for the molecule to be specifically drawn to the β-cleft of hemoglobin. Initial docking and energy minimization studies revealed that five different possibilities existed for the reagent to align itself within the β-cleft: (a) between β1-Lys-144 and β2-Val-1, (b) between β1-Lys-144 and β2-Lys-82, (c) between β1-Lys-144 and β2-Lys-132, (d) between β1-Lys-82 and β2-Lys-82, and (e) between β1-Lys-82 and β2-Val-1. The energy minimization and molecular dynamics simulation studies performed individually on these five different arrangements of the m-BCCEP-Hb complex suggested that the best fit with least energy and maximum reagent-Hb interactions (see below) occurs when m-BCCEP is aligned between Lys-144 on the β1 subunit and midway to N-terminus Val-1 and Lys-82 on the β2 subunit (intermediate between arrangements a and b described above). All atoms that were 12 Å or farther from the ligand were fixed with a temperature constant of 300 K. No constraints were applied to the remaining residues in and around the ligand site.

Hemoglobin Cross-Linking Reagent for β-Cleft Modification

Bioconjugate Chem., Vol. 21, No. 8, 2010 1499

The m-BCCEP-Hb complex was then energy minimized to convergence using consecutive steepest descent, conjugate gradient, VAO9A, and Newton energy minimization protocols. A stereoview of the energy minimized m-BCCEP-HbA0 complex is shown in Figure 3, while a close-up view of the same is given in Figure 4. Figure 5 highlights the concerned amino acid residues (lysines or R-terminus valine) whose appropriate (εor R-) amino groups fall within an approximate 9 Å distance from the two cross-linking sites of the reagent. The stereoview in Figure 3 clearly shows that the reagent m-BCCEP fits snugly into the β-cleft (DPG pocket) of hemoglobin. The close-up views in Figures 4 and 5 suggest that there are a total of four amino acid residues whose potentially crosslinkable amino groups lie within a ∼9 Å distance from the two activated ester carbonyl groups of the reagent, the anticipated cross-linking sites. Taking into account the sum of the van der Waal’s radii of the two concerned atoms involved in crosslinking [(C ) 2.0 + N ) 1.5) ) 3.5 Å total], coupled with the fact that the amino acids lining the periphery of the β-cleft are known to extend as much as 3-5 Å into the cleft to effect crosslinks (91), the consideration of all amino acids lying within a ∼9.0 Å radius from the reagent sites as potential candidates for cross-linking is reasonable. A femtosecond molecular dynamics simulation (300 K, 5000 iterations) of the above energy-minimized m-BCCEP-Hb complex in the preferred arrangement, soaked with an aqueous layer of 5 Å thickness all around, showed the concerned lysine residues on each β-subunit, lying in close proximity to the crosslinking sites of m-BCCEP. The range of distance between the appropriate ester carbonyl of m-BCCEP and the ε-amino group of lysine residues or the R-amino group of N-terminus valine, as computed from the graphs derived from dynamics trajectories, indicated that the closest contacts of β2-Val-1 (4.21 Å), β2-Lys82 (5.28 Å), and β1-Lys-144 (3.51 Å) are all well within the combined van der Waals radii of the two bonding atoms of the

concerned amino and carbonyl functions of Hb and m-BCCEP, respectively. Thus, the most preferred amino acids to participate in cross-linking by m-BCCEP appeared to be Val-1 and Lys82 on the β2 subunit and Lys-144 on the β1 subunit, which correspond to arrangements (a) and (b) described above. While the latter two are the suggested preferred arrangements, our simulation studies did not totally rule out the arrangements (c), (d), and (e) described earlier, which would require the involvement of an additional amino acid on each of the β-chains: Lys132 on β2 and Lys-82 on β1. We did, however, find some experimental evidence, based on MALDI MS analyses of tryptic digests of the cross-linked product (see below), supporting the arrangement (e). Synthesis of the Reagent m-BCCEP (1). Synthesis of the target reagent 1 involved four steps that are outlined in Scheme 1. The reaction of 3-hydroxybenzoic acid with N,N-dimethylformamide di-t-butyl acetal gave t-butyl 3-hydroxybenzoate (2). Condensation of (2) with acryloyl chloride in the presence of potassium t-butoxide in THF gave 3-t-butoxycarbonylphenyl acrylate (3). Next, 2 equiv of (3) were condensed with ammonium salt of hypophosphorous acid (59) with the silylating reagent N,O-bistrimethylsilylacetamide in anhydrous dichloromethane at 0 °C to produce bis[2-(3-t-butoxycarbonyl)phenoxycarbonylethyl]phosphinic acid (4). Treatment of the latter with trifluoroacetic acid resulted in removal of the tert-butyl protecting group to yield the target bis[2-(3-carboxyphenoxy)carbonylethyl]phosphinic acid (m-BCCEP; 1). The reagent 1 was converted to its trisodium salt by ion-exchange chromatography over AG 50W-X8 resin (Na+ form), followed by lyophilization. The crystalline trisodium salt is soluble in water and any aqueous buffers, and was directly used for reacting with hemoglobin. Cross-Linking Reaction of m-BCCEP with Hemoglobin. The cross-linking reaction of hemoglobin with the trisodium salt of m-BCCEP was carried out under oxygenated conditions in a

Table 1. Comparison of the Observed and Calculated Average Masses of Tryptic Peptides of Human Hb β-Chaina fragment residues (1-8) (9-17) (18-30) (31-40) (41-59) (60-61) (62-65) (66-66) (66-82) (67-82) (83-95) (96-104) (96-120)-S-S- (105-120) (105-120) (105-132) (121-132) (133-144) (133-146) (145-146)

calcd avg mass [MH]+

obsd avg mass [MH]+

tentative structural analysis of tryptic fragment

953.0 933.0 1867 1315.4 1337.4 1275.5 1298.5 2060.3 2082.3 246.3 412.5 147.2 1799.1 1837.1 1670.9 1422.6 1127.2 4569.4 1721.1 1799.1 3125.7 1379.6 1401.6 1150.4 1510.7 319.3

953.9 n.o.b 1867.8 1315.1 1337.3 1277.5 1298.1 2060.7 2082.1 n.a.a.b n.a.a.b n.a.a.b 1799.8 1837.8 1671.7 n.o.b 1128.4 4569.2 n.o.b 1799.8 3125.9 1379.6 1401.6 1150.1 1510.5 n.a.a.b

T1 n.o.b T2 (T1 + T2) - H2O T3 (T3 -1) + Na T4 (T4 - 1) + Na T5 (T5 - 1) + Na n.a.a.b (T6) n.a.a.b (T7) n.a.a.b (T8) (T8 + T9) - H2O (T8 + T9 - H2O - 1) + K T9 n.o.b (T10) T11 (T11 + T12 - H2O - 1)-S-S-(T12 - 1) + (-1 + Na) n.o.b (T12) (T12) + (-2 + 2K) (T12 + T13 - H2O) + (-2 + 2Na) T13 (T13 - 1) + Na T14 (T14 + T15 - H2O) + (-2 + Na + K) n.a.a.b (T15)

a The average mass of peptide fragments was calculated using the computer program, GPMAW (General Protein Mass Analyzer for Windows), v 2.0, available from Light House Data, Engvej 35, DK-5230, Odense M, Denmark. The following are the calculated masses for the peptide fragments resulting from tryptic digestion of the β-chain of hemoglobin and were employed in all subsequent computations involving these fragments: T1 (1-8), 952.08; T2 (9-17), 932.09; T3 (18-30), 1314.42; T4 (31-40), 1274.53; T5 (41-59), 2059.28; T6 (60-61), 245.33; T7 (62-65), 411.46; T8 (66-66), 146.19; T9 (67-82), 1669.90; T10 (83-95), 1421.59; T11 (96-104), 1126.24; T12 (105-120), 1720.11; T13 (121-132), 1378.55; T14 (133-144), 1149.36; T15 (145-146), 318.34. b The following are the abbreviations used in the Table: n.a.a., no assignment attempted; n.o., not observed; T, tryptic peptide.

1500 Bioconjugate Chem., Vol. 21, No. 8, 2010

Cai et al.

Scheme 1. Synthesis of the Cross-Linking Reagent m-BCCP and Its Trisodium Salt

phosphate buffer (pH 7.4) at 30 °C for 12 h using 2 mM human SFHb stock solution. Various concentrations (5×, 10×, 25×, 50× excess) of m-BCCEP (trisodium salt) were analyzed for its cross-linking capabilities. The cross-linking efficiency steadily increased with more cross-linker added and plateaued at 25fold excess of m-BCCEP as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analyses (Figure 6). Under extreme denaturing conditions of SDS-PAGE, crosslinked Hb is expected to migrate at nominal molecular weight of 32 kDa, and un-cross-linked Hb migrates at 16 kDa. In Figure 6, lane 1 is the SDS molecular weight markers, lane 2 is SFHb used as control, which showed predominantly a 16 kDa band representing monomeric, unmodified globin chains. Lanes 3 to 6 are the reaction mixtures, which showed strong bands at 32 kDa. Lane 5 (25×) and lane 6 (50×) have the best cross-linking effect, representing a substantial fraction (∼40%) of cross-linked Hb. Therefore, the best cross-linking efficiency was obtained with a 0.5 mM SFHb and a 25-fold excess of the reagent. To determine whether m-BCCEP cross-links between the R-R or β-β globin chains, the reaction mixture was analyzed by reverse-phase HPLC using a Vydac C4 column monitored at λmax 214 nm (see Figure 7). As indicated in Figure 7b, the reagent selectively modified the β globin chains, while the R chains were unaffected when compared to the spectrum of the

unreacted SFHb (Figure 7a). This was further confirmed by spiking the reaction mixture with SFHb (Figure 7c), when the unmodified β globin chain intensity increased. The individual R, β, and the cross-linked β-β chain fractions were further analyzed by MALDI-TOF, which confirmed that

Figure 6. SDS-PAGE analysis for SFHb and m-BCCEP-modified Hb.

Hemoglobin Cross-Linking Reagent for β-Cleft Modification

Bioconjugate Chem., Vol. 21, No. 8, 2010 1501

Figure 7. Reverse-phase (C4) HPLC chromatogram of (a) SFHb, (b) the cross-linked product with the reagent m-BCCEP (trisodium salt), and (c) the cross-linked product in (b) spiked with additional SFHb monitored at λmax 214 nm.

the masses of both R and β chains are ∼16 kDa, and only the cross-linked β-β chain has an additional band corresponding to a mass of ∼32 kDa. Tryptic Digestion. To determine the reactive sites of the cross-links, tryptic digestion analyses were carried out. Trypsin cleaves C-terminal lysine and arginine residues. The un-crosslinked β chain and the cross-linked β1-XL-β2 chain were separated by reverse-phase HPLC on a Vydac C4 column monitored at λmax 214 nm; the collected fractions were concentrated using Millipore Biomax 10 concentrators and then lyophilized into powder. Tryptic digestion was carried out for 24 h at 37 °C with gentle agitation. The digested globin chains

were analyzed by reverse-phase HPLC using a C18 Phenomenex column. The fragments obtained were collected from the HPLC column (Figure 8) and lyophilized individually. The tryptic digestion of the normal human Hb β-chain (containing 11 lysine and 3 arginine residues) is expected to produce 15 peptides (57), (92). MALDI-MS Analyses. The tryptic digestion fragments were analyzed by the matrix-assisted laser desorption/ionization timeof-flight mass spectrometry (MALDI-TOF-MS) (79-81), using the instrument specified under Materials and Methods above. Listed in Tables 1 and 2 are the observed average masses (MH+ ions) of predominant peptide fragments of Figure 8a,b, respec-

1502 Bioconjugate Chem., Vol. 21, No. 8, 2010

Cai et al.

Figure 8. The HPLC profiles of the tryptic digests of (a) pure β chain and (b) cross-linked β1-XL-β2 chain. The peptides were separated on a reverse-phase C18 column, using a linear gradient of CH3CN-H2O, containing 0.1% CF3CO2H.

tively, along with the calculated average masses and tentative structural analyses of each of those fragments. The presence of Na+ and K+ ions in a number of the observed fragments is believed to be from the buffer solution used during the crosslinking (57). Also listed in both tables are disulfide-linked peptides in the native as well as cross-linked hemoglobins. The exact sites involved in the cross-linking were determined by identification of ions whose masses were equal to the sum of two tryptic fragments plus an additional mass equal to that of the cross-linking reagent minus the leaving group moieties (in this case, a mass of 176.1 corresponds to the m-BCCEP tether). The observed ions listed in Table 2 suggest that there are three intersubunit cross-links in the hemoglobin modified by m-BCCEP (Figure 9). They are as follows: (a) β2 Val 1 to β1 Lys 144, supported by ions at m/z 2466.45, 6462.95; (b) β2 Lys 82 to β1 Lys 144, validated by ions at m/z 2996.22, 4593.50, and 4781.44; and (c) β2 Val 1 to β1 Lys 82, as corroborated by ions at m/z 4587.14 and 5535.7. The observed third crosslink in (c) was somewhat unexpected based on our molecular modeling results, which showed that the distance between the carbonyl carbon of the reagent and the ε-amino group of β1lysine is ∼12 Å, a bit farther away for effective cross-linking. However, considering that the described molecular dynamics simulations studies did not completely rule out the possibility of the reagent aligning itself between β1-lysine and β2-valine (arrangement e described under Molecular Modeling Studies), this result is not totally surprising. It is also to be noted that since cross-linking is anticipated to block the action of trypsin

at the lysine (or arginine) residue involved in the cross-link, all the cross-linked peptide fragments might contain additional peptide units beyond the point of cross-link. The computation of molecular weight of each cross-linked peptide species is based on (a) adding the molecular weights of the concerned peptides; (b) subtraction of 2 mass units and addition of 176.1 mass units to account for the addition of the reagent tether to two peptides; (c) subtraction of 18 mass units corresponding to a molecule of water for each pair of peptides present; for example, 18 (1 H2O) or 36 (2 H2O) mass units would be deducted from a fragment containing 2 or 3 peptide units, respectively; and (d) one mass unit is deducted for each sodium or potassium ion added. As an example, an ion at m/z 5535.7, representing a cross-link between β1 lysine 1 and β2 lysine 144, is computationally reconciled as follows: [(β1T1 ) 952.1) + (β2T8 + β2T9 + β2T10 + β2T11 - 3H2O ) 4313.9) + (XL ) 176.1) - (peptide amino hydrogens ) 2) + (Na + 2K - 3 ) 98) + (protonation of the cross-linked peptide during the ionization process ) 1) ) total m/z 5535.7]. All other ions listed in Table 2 can be similarly accounted for as shown in the last column of the table. Some peptide fragments listed in Table 2 also suggest that the peptides contained fragments resulting from m-BCCEPlinked single β-chain, due to being hydrolyzed before m-BCCEP reaction with another amino group on the other β-chain (59). Oxygen Equilibrium (P50) Studies. The m-BCCEP-modified hemoglobin product was analyzed for oxygen affinity (P50) characteristics at 37 °C. For the sake of comparison, P50 values

Hemoglobin Cross-Linking Reagent for β-Cleft Modification

Bioconjugate Chem., Vol. 21, No. 8, 2010 1503

Table 2. Comparison of the Observed and Calculated Average Masses of Tryptic Peptides of Human Hb β-Chain Cross-Linked with the Reagent m-BCCEPa fragment residues

calcd avg mass [M+H]+

obsd avg mass [M+H]+

tentative structural analysis of tryptic fragment

(1-8) XL (66-104) (1-8) XL (66-95) + Reagent (1-8) XL (133-144) (1-40) XL (133-146) + Reagent

953.0 975.1 1035.0 1051.0 1067.0 5535.1 4586.9 2466.5 6463.8

953.46 975.50 1035.06 1051.09 1067.04 5535.70 4587.14 2466.45 6462.95

933.0 1315.4 1275.5 2060.3 246.3 412.5 2164.5 2230.6 2274.6 147.2 1837.1 4593.2 4783.5 3312.7 1670.9 2994.4 1422.6 1532.6 2688.8 2810.9 1127.2 1171.2 1721.1 1831.1 1379.6 1150.4 1172.4 1194.4 1364.5 1450.7 1642.8 1664.8 319.3

933.17 1315.48 1275.01 2060.15 n.a.a.b n.a.a.b 2164.31 2229.70 2274.76 n.a.a.b 1837.61 4593.50 4781.44 3312.78 n.o.b 2996.22 n.o.b 1532.59 2689.57 2810.62 n.o.b 1171.94 n.o.b 1831.31 1379.42 n.o.b 1172.91 1194.87 1364.31 1450.68 1642.84 1664.69 n.a.a.b

T1 (T1 - 1) + Na (T1) + (-3 + 2Na + K) (T1) + (-3 + Na + 2K) (T1) + (-3 + 3K) (T1 - 1) XL (T8 + T9 + T10 + T11 - 3H2O - 1) + (-3 + Na + 2K) (T1 - 1) XL (T8 + T9 + T10 - 2H2O - 1) + (-1 + Rb) + (-3 + 3Na) (T1 - 1) XL (T14 - 1) + (-5 + 5K) (T1 + T2 + T3 + T4 - 3H2O - 1) XL (T14 + 15 - H2O - 1) + (-1 + Rb) + (-6 + 6K) T2 T3 T4 T5 n.a.a.b (T6) n.a.a.b (T7) (T7 + T8 - H2O - 1) XL (T14 + T15 - H2O - 1) (T7 + T8 + T9 - 2H2O - 1) + K (T7 + T8 + T9 - 2H2O) + (-3 + 2Na + K) n.a.a.b (T8) (T8 + T9 - H2O - 1) + K (T8 + T9 - H2O - 1) XL (T13 + T14 - H2O - 1) + (-5 + 5Na) (T8 + T9 - H2O - 1) XL (T13 + T14 + T15 - 2H2O - 1) (T8 + T9 + T10 - 2H2O) + (-5 + 5Na) n.o.b (T9) (T9 - 1) XL (T14 - 1) n.o.b (T10) (T10) + (-5 + 5Na) (T10 + T11 - H2O) + (-5 + 2Na + 3K) (T10 + T11 - H2O - 1) + Rb + (-4 + 4Na) n.o.b (T11) (T11) + (-2 + 2Na) n.o.b (T12) (T12) + (-5 + 5Na) T13 n.o.b (T14) (T14 - 1) + Na (T14) + (-2 + 2Na) (T14 - 1) + Rb + (-1 + Na) T14 + T15 - H2O (T14 + T15 - H2O - 1) + Rb (T14 + T15 - H2O - 1) + Rb+ (-1 + Na) n.a.a.b (T15)

(1-8)

(9-17) (18-30) (31-40) (41-59) (60-61) (62-65) (62-66) XL (133-146) (62-82) (66-66) (66-82) (66-82) XL (121-144) (66-82) XL (121-146) (66-95) (67-82) (67-82) XL (133-144) (83-95) (83-104) (83-104) + Reagent (96-104) (105-120) (121-132) (133-144) (133-144) + Reagent (133-146) (133-146) + Reagent (145-146)

a The average mass of peptide fragments was calculated using the computer program, GPMAW (General Protein Mass Analyzer for Windows), v 2.0, available from Light House Data, Engvej 35, DK-5230, Odense M, Denmark. The following are the calculated masses for the peptide fragments resulting from tryptic digestion of the β-chain of hemoglobin and were employed in all subsequent computations involving these fragments: T1 (1-8), 952.08; T2 (9-17), 932.09; T3 (18-30), 1314.42; T4 (31-40), 1274.53; T5 (41-59), 2059.28; T6 (60-61), 245.33; T7 (62-65), 411.46; T8 (66-66), 146.19; T9 (67-82), 1669.90; T10 (83-95), 1421.59; T11 (96-104), 1126.24; T12 (105-120), 1720.11; T13 (121-132), 1378.55; T14 (133-144), 1149.36; T15 (145 146), 318.34. b The following are the abbreviations used in the Table: n.a.a., no assignment attempted; n.o., not observed; T, tryptic peptide; XL ) C(O)(CH2)2P(O)(OH)(CH2)2C(O) ) cross-linked reagent with a mass ) 176.1; R ) C(O)(CH2)2P(O)(OH)(CH2)2CO2H ) reagent reacted on only one side, while the other side is hydrolyzed, with a mass of 193.12. Here, reagent is m-BCCEP (trisodium salt).

Figure 9. Three intersubunit cross-links present in the m-BCCEP modified Hb. The one with a broken arrow is a minor cross-link.

were also measured under the same conditions for the un-crosslinked, stroma-free Hb (SFHb). The graphs of PO2 vs % Hb saturation for SFHb and the BCCEP-modified Hb are shown in Figure 10. The computed P50 value for the modified Hb was 25.8 Torr, in contrast to that of the stroma-free Hb (14.19 Torr). Thus, the m-BCCEP-modified Hb showed significantly lowered oxygen affinity as compared with the stroma-free hemoglobin. This was further attested by the observed right-shifting of O2 equilibrium curve of the cross-linked product relative to that of SFHb. Furthermore, the reasonable retainment of oxygen

binding cooperativity after the cross-link formation was suggested by the Hill coefficient n (cooperativity of oxygen binding is measured by the Hill coefficient, n, which is computed by the Hill equation, Y ) (PO2)n/[(P50)n + (PO2)n], where Y is the fraction of Hb saturated. The value of n in whole blood ranges 2.3-2.5), generated from the Hill plots of P50 values, which came out to 1.91. This value in whole blood ranges 2.0-2.5 (93). However, while the Hill coefficient would be a quantitative indicator of oxygen-binding cooperativity, the computed value from such plots for the m-BCCEP-modified Hb is, nonetheless, less meaningful in view of the anticipated and observed multiple cross-linking sites and the consequent heterogeneity of the m-BCCEP-modified Hb.

CONCLUSION The design and synthesis of an efficacious hemoglobin crosslinking reagent, employing modern molecular modeling and organic synthesis techniques, have been reported. The title reagent m-BCCEP can be synthesized in four facile, convenient, and good-yielding steps, employing readily available, inexpen-

1504 Bioconjugate Chem., Vol. 21, No. 8, 2010

Cai et al.

Figure 10. Oxygen equilibrium curves for (a) stroma-free Hb and (b) m-BCCEP-modified Hb.

sive starting materials, and can be easily adapted to high-scale production. It has an indefinite shelf life and can be converted into its crystalline trisodium salt that can be employed to crosslink hemoglobin under aqueous biological medium and ambient oxygenated reaction conditions, requiring no special precautions to either preclude oxygen or include carbon monoxide as is the case with many known hemoglobin cross-linking agents. The tryptic digestion and MALDI mass spectral analyses of the modified β-chain fragments suggest that the reagent predominantly forms two cross-links between β1 Lys 144 and either β2 Val 1 or β2 Lys 82, although a minor third cross-link was also found between β1 Lys 82 and β2 Val 1. These results were remarkably consistent with those predicted by molecular modeling studies of the m-BCCEP-Hb A0 complex. The oxygen affinity of the cross-linked product is considerably lower (P50 ) 25.8 Torr) as desired in comparison with that of stroma-free Hb (P50 ) 14.2 Torr) and is somewhat comparable to the oxygen affinity of the whole blood (P50 ) 27 Torr). The cross-linked product also retains some of the oxygen-binding cooperativity characteristics of the native protein, as revealed by its Hill coefficient value (n ) 1.91), which is significantly higher than that of the previously reported BPPCEP-modified hemoglobin products, all of which had the n value equal to or less than 1.0 (59). Finally, the field of blood substitutes based on cell-free hemoglobin has a vast and rich history with a wide variety of cross-linking reagents, modified Hb products derived

therefrom, and data on their oxygen affinity characteristics, osmotic pressure, vasoconstriction, and other physiological properties, as well as on a few products that have made to the clinical trials. In that context, discovering another Hbcross-linking reagent with novel characteristics will only be a first step toward an optimal, physiologically feasible, and clinically compatible blood substitute. If history is any indication, even the most ideally modified Hb products are likely to suffer from one or the other historical problems of short circulation times in the bloodstream, inadequate colloidal osmotic pressures (COP), or vasoconstriction due to scavenging of nitric oxide, the endothelium-derived relaxation factor (46-50). However, these past setbacks should, nonetheless, discourage or halt future scientific efforts to realize an eventually optimal blood substitute that has so much stake in success outcome of almost every major field of modern medicine. To alleviate some of the problems currently encountered in blood substitute field, it may be necessary to increase the steric bulk of the cross-linked hemoglobin through polymerization. This will allow its retention in circulation for prolonged periods of time (44, 45). The increased size would also prevent its facile sieving through the endothelial lining of blood vessels, where they are known to scavenge nitric oxide. Such an effort based on m-BCCEP framework and other novel organic reagents is currently in progress in this laboratory.

Hemoglobin Cross-Linking Reagent for β-Cleft Modification

ACKNOWLEDGMENT This research was supported by a grant (0755372U) from the American Heart Association. We thank Dr. Victor Macdonald, Chief, Division of Blood Products & Storage, Walter Reed Army Institute of Research, Washington, DC, for generous supply of stroma-free hemoglobin employed in our studies. We are highly grateful to Dr. Fantao Meng of Albert Einstein College of Medicine for assistance in our oxygen equilibrium studies on modified hemoglobins.

LITERATURE CITED (1) Scholer, M., Frietsch, T., Jambor, C., and Knels, R. (2010) Artificial blood--coming soon or never reaching clinical maturity? Dtsch. Med. Wochenschr. 135, 575–581. (2) Winslow, R. M. (1992) Hemoglobin-based Red Cell Substitutes, The Johns Hopkins University Press, Baltimore and London. (3) Winslow, R. M. (2003) Current status of blood substitute research: towards a new paradigm. J. Intern. Med. 253, 508– 517. (4) Winslow, R. M. (2006) Current status of oxygen carriers (‘blood substitutes’). Vox Sang. 91, 102–10. (5) Winslow, R. M., Vandegriff, K. D., and Intaglietta, M. (1996) Blood Substitutes: New Challanges (Winslow, R. M., Vandegriff, K. D., and Intaglietta, M., Eds) pp 61-63, Birkhauser, Boston. (6) Alayash, A. I. (2001) Is there a need for blood substitutes in the new millennium and what should we expect in the way of safety and efficacy? A research perspective. Artif. Cells Blood Substit. Immobil. Biotechnol. 29, 1–3. (7) Chang, T. M. (2004) Hemoglobin-based red blood cell substitutes. Artif. Organs 28, 789–94. (8) Jahr, J. S., Nesargi, S. B., Lewis, K., and Johnson, C. (2002) Blood substitutes and oxygen therapeutics: an overview and current status. Am. J. Ther. 9, 437–43. (9) Klein, H. G. (2002) Blood substitutes: how close to a solution? Oncology (Williston Park) 16, 147–51. (10) Squires, J. E. (2002) Artificial blood. Science 295, 1002–5. (11) McCullough, J. J. (1998) Transfusion Medicine, pp 91-93, McGraw-Hill, New York. (12) Rossi, E. C., Simon, T. L., and Moss, G. S. (1991) Principles of transfusion medicine, pp 1-11, Williams & Wilkins, Baltimore. (13) Fronticelli, C., and Koehler, R. C. (2009) Design of recombinant hemoglobins for use in transfusion fluids. Crit. Care Clin. 25, 357–371. (14) Creteur, J., and Vincent, J.-L. (2009) Potential uses of hemoglobin-based oxygen carriers in critical care medicine. Crit. Care Clin. 25, 311–324. (15) Winslow, R. M. (2007) Red cell substitutes. Semin. Hematol. 44, 51–9. (16) Greenburg, A. G. (2009) The ideal blood substitute. Crit. Care Clin. 25, 415–424. (17) Baron, J. F., and Treib, J. (1998) Volume Replacement, pp 1-143, Springer, Berlin. (18) Lowe, K. C. (1998) Fluorocarbon Emulsions as Blood Substitutes, in Blood Substitutes Present and Future PerspectiVes (Tsuchida, E., Ed.) pp 327-338, Elsevier Science S.A., Switzerland. (19) Riess, J. G. (1993) Overview of Progress in the Fluorocarbon Approach to in ViVo Oxygen Delivery, in Blood Substitutes and Oxygen Carriers (Chang, T. M. S., Ed.) pp 24-43, Marcel Dekker, Inc, New York. (20) Cohn, C. S., and Cushing, M. M. (2009) Oxygen therapeutics: perfluorocarbons and blood substitute safety. Crit. Care Clin. 25, 399–414. (21) Farmer, M. C, and Gaber, B. P. (1987) Liposome-encapsulated hemoglobin as an artificial oxygen carrying system. Methods Enzymol. 149, 184–200. (22) Rohlfs, R. J., and Vandegriff, K. D. (1996) Non-Phospholipid Liposomes: A Novel Method for the Preparation of Hemoglobin

Bioconjugate Chem., Vol. 21, No. 8, 2010 1505 Containing Lipid Vesicles, in Blood Substitutes: New Challenges (Winslow, R. M., Vandegriff, K. D., and Intaglietta, M., Eds.) pp163-184, Birkha¨user, Boston. (23) Rudolph, A. S. (1995) Encapsulation of Hemoglobin in Liposomes, in Blood Substitutes: Physiological Basis of Efficacy (Winslow, R. M., Vandegriff, K. D., and Intaglietta, M., Ed.) pp 90-104, Birkhauser, New York. (24) Chang, T. M. S. (1989) Red Blood Cell Substitutes: Microencapsulated Hemoglobin and Cross-Linked Hemoglobin Including Pyridoxylated Polyhemoglobin & Conjugated Hemoglobin, in Blood Substitutes (Chang, T. M. S., and Geyer, R. P., Ed.) pp 11-29, Marcel Dekker, Inc., New York. (25) Alayash, A. I. (1998) Hemoglobin-based Blood Substitutes and Mechanisms of Toxicity, in Blood Substitutes Present and Future PerspectiVes (Tsuchida, E., Ed.) pp 201-210, Elsevier Sciences S.A., Switzerland. (26) Alayash, A. I. (1999) Hemoglobin-based blood substitutes: oxygen carriers, pressor agents, or oxidants? Nat. Biotechnol. 17, 545–9. (27) Alayash, A. I. (2000) Hemoglobin-based blood substitutes and the hazards of blood radicals. Free Radic. Res. 33, 341–8. (28) Eastman, A. L., and Minei, J. P. (2009) Comparison of hemoglobin-based oxygen carriers to stored human red blood cells. Crit. Care Clin. 25, 303–310. (29) Napolitano Lena, M. (2009) Hemoglobin-based oxygen carriers: first, second or third generation? Human or bovine? Where are we now? Critical Care Clinics 25, 279–301, Table of Contents. (30) Rudolph, A. S. (1998) Hemoglobin-Based Blood Substitutes, in Frontiers in Tissue Engineering (Patrick, C. W., Jr., Mikos, A. G., and McIntyre, L. V., Eds.) pp481-494, Elsevier, New York. (31) Maton, A., Hopkins, J., McLaughlin, C. W., Johnson, S., Warner, M. Q., LaHart, D., and Wright, J. D. (1993) Human Biology and Health, Pearson Prentice Hall, Englewood Cliffs, NJ. (32) Amberson, W. (1937) Blood substitute. Biol. ReV. 12–48. (33) Rabiner, S. F., Helbert, J. R., Lopas, H., and Friedman, L. H. (1967) Evaluation of a stroma-free hemoglobin solution for use as a plasma expander. J. Exp. Med. 126, 1127–42. (34) Savitsky, J. P., Doczi, J., Black, J., and Arnold, J. D. (1978) A clinical safety trial of stroma-free hemoglobin. Clin. Pharmacol. Ther. 23, 73–80. (35) Horton, H., Moran, L., Ochs, R., Rawn, J., and Scrimgeour, K. (1993) Principles of Biochemistry, Neil Patterson, New Jersey. (36) Astrup, P., and Rorth, M. (1971) Oxygen Affinity of Hemoglobin and Red Cell Acid Base Status, Academic Press, New York. (37) Haney, C. R., Buehler, P. W., and Gulati, A. (2000) Purification and chemical modifications of hemoglobin in developing hemoglobin based oxygen carriers. AdV. Drug DeliVery ReV. 40, 153–69. (38) Viele, M. K., Weiskopf, R. B., and Fisher, D. (1997) Recombinant human hemoglobin does not affect renal function in humans: analysis of safety and pharmacokinetics. Anesthesiology 86, 848–58. (39) Arnone, A. (1972) X-ray diffraction of binding of 2,3diphosphoglycerate to human dexoyhaemoglobin. Nature 237, 146–149. (40) Bakker, J. C., Berbers, G. A., Bleeker, W. K., den Boer, P. J., and Biessels, P. T. (1992) Preparation and characterization of crosslinked and polymerized hemoglobin solutions. Biomater. Artif. Cells Immobilization Biotechnol. 20, 233–41. (41) Gould, S. A., Sehgal, L. R., Sehgal, H. L., and Moss, G. S. (1992) Artificial blood: current status of hemoglobin solutions. Crit. Care Clin. 8, 293–309. (42) Spahn, D. R. (2000) Current status of artificial oxygen carriers. AdV. Drug DeliVery ReV. 40, 143–51. (43) Bolin, R., Geyer, R., and Nemo, G. E. (1983) AdVances in Blood Substitute Research, pp 1-468, Alan R. Liss, Inc., New York.

1506 Bioconjugate Chem., Vol. 21, No. 8, 2010 (44) Jahr, J. S., Moallempour, M., and Lim, J. C. (2008) HBOC201, hemoglobin glutamer-250 (bovine), Hemopure (Biopure Corporation). Expert Opin. Biol. Ther. 8, 1425–1433. (45) Napolitano, L. M. (2009) Hemoglobin-based oxygen carriers: first, second or third generation? Human or bovine? Where are we now? Crit. Care Clin. 25, 279–301. (46) Allen, B. W., and Piantadosi, C. A. (2006) How do red blood cells cause hypoxic vasodilation? The SNO-hemoglobin paradigm. Am. J. Physiol. 291, H1507–H1512. (47) Angelo, M., Hausladen, A., Singel, D. J., and Stamler, J. S. (2008) Interactions of NO with hemoglobin: from microbes to man. Methods Enzymol. 436, 131–168. (48) Babbar, R., and Agarwal, S. (2005) Nitric oxide-hemoglobin interactions: Role in oxygen uptake and delivery. Ind. J. Physiol. Pharmacol. 49, 378–382. (49) Goedecke, A. (2006) On the impact of NO-globin interactions in the cardiovascular system. CardioVasc. Res. 69, 309–317. (50) von Dobschuetz, E., Hutter, J., Hoffmann, T., and Messmer, K. (2004) Influence of hemoglobin-based oxygen carriers with different nitric oxide scavenging properties on the pancreatic microcirculation after hemorrhagic shock. Prog. Appl. Microcirc. 25, 86–94. (51) Alleyne, T. A., Wilson, M. T., Antonini, G., Malatesta, F., Vallone, B., Sarti, P., and Brunori, M. (1992) Investigation of the electron-transfer properties of cytochrome c oxidase covalently cross-linked to Fe- or Zn-containing cytochrome c. Biochem. J. 287, 951–6. (52) Hosmane, R. S. (1984) Imidates in organic synthesis: Methyl N-cyanomethylmethanimidate. Tetrahedron Lett. 25, 363–6. (53) Hosmane, R. S., and Bertha, C. (1988) A novel bifunctional reagent: 2,2′-bis(methoxymethylene)-2,2′-sulfonyldiacetonitrile. Tetrahedron Lett. 29, 5847–50. (54) Hosmane, R. S., Bertha, C., Siriwardane, U., and Hosmane, N. (1990) Reagents for bioorganic synthesis. 5. the synthesis of two potential cross-linking reagents: 2,2′-sulfonylbis[3-(benzyamino)-(E,E)-N-(2-oxoethyl)propenamide](SBBOP) and 2,2′sulfonylbis[3-(benzylamino)-(E,E)-N(-2-chloroethyl)propenamide] (SBBCP). J. Org. Chem. 55, 5206–16. (55) Hosmane, R. S., and Bertha, C. M. (1990) A versatile, highly reactive, cross-linking reagent: 2,2′-sulfonylbis[3-methoxy-(E,E)2-propenenitrile]. Biochem. Biophys. Res. Commun. 166, 567– 73. (56) Hosmane, R. S., Burnett, F., and Albert, M. (1984) Reagents for bioorganic synthesis: preparation, properties, and reactions of ethyl and methyl N-(cyanomethyl)methanimidates. J. Org. Chem. 49, 1212–5. (57) Hosmane, R. S., Peri, S. P., Bhadti, V. S., and Macdonald, V. W. (1998) Bis[2-(4-carboxyphenoxy)carbonylethyl]phosphinic acid (BCCEP): a novel affinity reagent for the beta-cleft modification of human hemoglobin. Bioorg. Med. Chem. 6, 767– 83. (58) Mayer, A. N., and Barany, F. (1995) Photoaffinity crosslinking of TaqI restriction endonuclease using an aryl azide linked to the phosphate backbone. Gene 153, 1–8. (59) Roach, T. A., Macdonald, V. W., and Hosmane, R. S. (2004) A novel site-directed affinity reagent for cross-linking human hemoglobin: bis[2-(4-phosphonooxyphenoxy)carbonylethyl]phosphinic acid. J. Med. Chem. 47, 5847–59. (60) Wohlfeil, E. R., and Hudson, R. A. (1991) Synthesis and characterization of a heterobifunctional mercurial cross-linking agent: incorporation into cobratoxin and interaction with the nicotinic acetylcholine receptor. Biochemistry 30, 7231–41. (61) Zhu, Q., and Courtney, R. J. (1994) Chemical cross-linking of virion envelope and tegument proteins of herpes simplex virus type 1. Virology 204, 590–9. (62) Bertha, C. M., Hosmane, S. H., Zhang, H., and Hosmane, N. S. (1994) Reagents for bioorganic synthesis. 6. The design and synthesis of two novel polyfunctional organic reagents for biomacromolecular modification. J. Biomol. Struct. Dyn. 12, 413– 37.

Cai et al. (63) Chang, T. M. S. (1993) Blood Substitutes and Oxygen Carriers, pp 1-871, Marcel Dekker Inc., Montreal. (64) Chang, T. M. S. (1997) Blood Substitutes: Principles, Methods, Products and Clinical Trials, pp 1-142, Vol 1, Karger Lander Systems. (65) Crapatureanu, S., Serbanescu, R., Brevitt, S. B., and Kluger, R. (1999) Molecular necklaces. Cross-linking hemoglobin with reagents containing covalently attached ligands. Bioconjugate Chem. 10, 1058–67. (66) Jones, R. T., Head, C. G., Fujita, T. S., Shih, D. T., Wodzinska, J., and Kluger, R. (1993) Modification of human hemoglobin with methyl acyl phosphates derived from dicarboxylic acids. Systematic relationships between cross-linked structure and oxygen-binding properties. Biochemistry 32, 215–23. (67) Jones, R. T., Shih, D. T., Fujita, T. S., Song, Y., Xiao, H., Head, C., and Kluger, R. (1996) A doubly cross-linked human hemoglobin. Effects of cross-links between different subunits. J. Biol. Chem. 271, 675–80. (68) Kluger, R. (1994) 1994 Syntex Award Lecture. Anionic electrophiles, protein modification, and artificial blood. Can. J. Chem. 72, 2193–2197. (69) Kluger, R., Jones, R. T., and Shih, D. T. (1994) Cross-linking hemoglobin by design: lessons from using molecular clamps. Artif. Cells Blood Substit. Immobil. Biotechnol. 22, 415–28. (70) Kluger, R., and Li, X. (1997) Efficient chemical introduction of a disulfide cross-link and conjugation site into human hemoglobin at beta-lysine-82 utilizing a bifunctional aminoacyl phosphate. Bioconjugate Chem. 8, 921–6. (71) Kluger, R., Lock-O’Brien, J., and Teytelboym, A. (1999) Connecting proteins by design. cross-linked bis-hemoglobin. J. Am. Chem. Soc. 121, 6780–6785. (72) Kluger, R., and Stefano, V. D. (2000) Mechanism of sitedirected protein cross-linking. protein-directed selectivity in reactions of hemoglobin with aryl trimesates. J. Org. Chem. 65, 214–9. (73) Peri, S. P., Bhadti, V. S., Hosmane, R. S., and Macdonald, V. W. (1997) Bis[2-(4-carboxyphenoxy)carbonylethyl]phosphinic acid (BCCEP): a new reagent for hemoglobin modification. Artif. Cells Blood Substit. Immobil. Biotechnol. 25, 511–20. (74) Peri, S. P., Bhadti, V. S., Somerville-Armstrong, K. S., Liang, R., Macdonald, V. W., and Hosmane, R. S. (1999) Affinity reagents for cross-linking hemoglobin: bis(phenoxycarbonylethyl)phosphinic acid (BPCEP) and bis(3-nitrophenoxycarbonylethyl)phosphinic acid (BNCEP). Hemoglobin 23, 1–20. (75) Ueno, H., Pospischil, M. A., Manning, J. M., and Kluger, R. (1986) Site-specific modification of hemoglobin by methyl acetyl phosphate. Arch. Biochem. Biophys. 244, 795–800. (76) Harrington, J. P., Gonzalez, Y., and Hirsch, R. E. (2000) Redox concerns in the use of acellular hemoglobin-based therapeutic oxygen carriers: the role of plasma components. Artif. Cells Blood Substit. Immobil. Biotechnol. 28, 477–92. (77) Everse, J., and Hsia, N. (1997) The toxicities of native and modified hemoglobins. Free Radic. Biol. Med. 22, 1075–99. (78) Silva, M. M., Rogers, P. H., and Arnone, A. (1992) A third quaternary structure of human hemoglobin A at 1.7A0. resolution. J. Biol. Chem. 267, 17248–17256. (79) Pimenova, T., Pereira, C. P., Schaer, D. J., and Zenobi, R. (2009) Characterization of high molecular weight multimeric states of human haptoglobin and hemoglobin-based oxygen carriers by high-mass MALDI MS. J. Sep. Sci. 32, 1224–1230. (80) Porubleva, L. V., Rebriiev, A. V., Gromovoy, T. Y., Minia, I. J., and Obolenskaya, M. Y. (2009) MALDI-TOF mass spectrometry in the investigation of high-molecular biological compounds. Ukrains’kii Biokhimichnii Zhurnal 81, 46–56. (81) Vestal, M. L. (2009) Modern MALDI time-of-flight mass spectrometry. J. Mass Spectrom. 44, 303–317. (82) Abraham, D. J., Safo, M. K., Boyiri, T., Danso-Danquah, R. E., Kister, J., and Poyart, C. (1995) How allosteric effectors can bind to the same protein residue and produce opposite shifts in the allosteric equilibrium. Biochemistry 34, 15006–20.

Hemoglobin Cross-Linking Reagent for β-Cleft Modification (83) Boyiri, T., Safo, M. K., Danso-Danquah, R. E., Kister, J., Poyart, C., and Abraham, D. J. (1995) Bisaldehyde allosteric effectors as molecular ratchets and probes. Biochemistry 34, 15021–36. (84) Randad, R. S., Mahran, M. A., Mehanna, A. S., and Abraham, D. J. (1991) Allosteric modifiers of hemoglobin. 1. Design, synthesis, testing, and structure-allosteric activity relationship of novel hemoglobin oxygen affinity decreasing agents. J. Med. Chem. 34, 752–7. (85) Wireko, F. C., Kellogg, G. E., and Abraham, D. J. (1991) Allosteric modifiers of hemoglobin. 2. Crystallographically determined binding sites and hydrophobic binding/interaction analysis of novel hemoglobin oxygen effectors. J. Med. Chem. 34, 758–67. (86) Trimble, S. P., Marquardt, D., and Anderson, D. C. (1997) Use of designed peptide linkers and recombinant hemoglobin mutants for drug delivery: in vitro release of an angiotensin II analog and kinetic modeling of delivery. Bioconjugate Chem. 8, 416–423. (87) Pechik, I., Ji, X., Fidelis, K., Karavitis, M., Moult, J., Brinigar, W. S., Fronticelli, C., and Gilliland, G. L. (1996) Crystallographic, molecular modeling, and biophysical characterization of the valineβ67 (E11) -> threonine variant of hemoglobin. Biochemistry 35, 1935–45.

Bioconjugate Chem., Vol. 21, No. 8, 2010 1507 (88) Wajcman, H., Kister, J., Prome, D., Blouquit, Y., Prehu, C., Poyart, C., and Galacteros, F. (1995) Interaction of 2 amino acid substitutions within the same beta chain of human hemoglobin: the examples of Hb Corbeil and Hb Villeparisis. C. R. Acad. Sci. Ser. III 318, 785–94. (89) Schaad, O., Vallone, B., and Edelstein, S. J. (1993) Critical residues responsible for self-association differences of hemoglobin Iˆ( and Iˆ2 chains: Analysis by molecular modeling. C. R. Acad. Sci. Ser. III 316, 564–71. (90) Cretegny, I., and Edelstein, S. J. (1993) Double strand packing in hemoglobin S fibers. J. Mol. Biol. 230, 733–8. (91) Arnone, A., Benesch, R. E., and Benesch, R. (1977) Structure of human deoxyhemoglobin specifically modified with pyridoxal compounds. J. Mol. Biol. 115, 627–642. (92) Yang, T., Horejsh, D. R., Mahan, K. J., Zaluzec, E. J., Watson, T. J., and Gage, D. A. (1996) Mapping cross-linking sites in modified proteins with mass spectrometry: an application to cross-linked hemoglobins. Anal. Biochem. 242, 55–63. (93) Miller, I. F. (1978) Synthetic blood substitutes: where are we and where do we go from here? CRC Crit. ReV. Bioeng. 3, 149–179. BC100113Y