Chemical Characterization of Diaspirin Cross-Linked Hemoglobin

A lack of specificity associated with chemical modification methods used in the preparation of certain hemoglobin (Hb)-based oxygen carriers (HBOCs) m...
1 downloads 0 Views 208KB Size
Download PDF
Anal. Chem. 2006, 78, 4634-4641

Chemical Characterization of Diaspirin Cross-Linked Hemoglobin Polymerized with Poly(ethylene glycol) Paul W. Buehler,† Robert A. Boykins,‡ Scott Norris,‡ and Abdu I. Alayash*,†

Laboratory of Biochemistry and Vascular Biology, Division of Hematology, and Laboratory of Biophysics, Division of Bacterial, Parasitic and Allergenic Products, Center for Biologics Evaluation and Research (CBER), Food and Drug Administration (FDA), Bethesda, Maryland 20892

A lack of specificity associated with chemical modification methods used in the preparation of certain hemoglobin (Hb)-based oxygen carriers (HBOCs) may alter Hb structure and function, as amino acids located in critical regions (e.g., r-β interfaces and the 2,3-DPG binding pocket) may unintentionally be targeted. Hb protein surface modifications with various poly(ethylene glycol) (PEG) derivatives have been used as conjugating and polymerizing agents with the intent of improving reaction site specificity/reproducibility and ultimately reducing the untoward hypertensive response due to nitric oxide scavenging by smaller molecular size tetrameric species (i.e., 64 kDa) in HBOC solutions. Previous experiments performed in our laboratory have evaluated the influence of polymerization of diaspirin r-r cross-linked Hb (rrDBBF-Hb) with a bifunctional modified PEG, bis(maleoylglycylamide) PEG (BMAA-PEG), in terms of oxygen carrying capacity, redox properties, hypertensive response, and renal clearance in rats. The data presented in this paper specifically evaluate the influence of BMAAPEG on rr-DBBF-Hb (Poly-rr-DBBF-Hb) to identify molecular weight distribution, protein conformation, and site-specific modification, as well as to provide insight into the previously determined in vitro and in vivo functional and vasoactive characteristics of this HBOC. Chemical analysis performed herein reveals nonspecific modifications induced by BMAA-PEG that result in the full modification of rr-DBBF-Hb leaving no tetrameric crosslinked starting material in solution. These data are inconsistent with the continuing assumption that molecular size (i.e., 64 kDa) has a direct influence on HBOC-mediated vasoactivity and that other protective strategies should be considered to control blood pressure imbalances. Modified poly(ethylene glycol) (PEG) is utilized primarily to enhance the circulation time of peptides and proteins, thus prolonging their therapeutic effect. PEGylation technology has also gained acceptance in the preparation of hemoglobin (Hb)based oxygen carriers (HBOCs) as both a random1,2 and a more * To whom correspondence should be addressed. Telephone: (301) 827-3813. Fax: (301) 435-4034. E-mail: [email protected]. † Laboratory of Biochemistry and Vascular Biology, Division of Hematology. ‡ Laboratory of Biophysics, Division of Bacterial, Parasitic and Allergenic Products.

4634 Analytical Chemistry, Vol. 78, No. 13, July 1, 2006

site-specific means3,4 of directing functional group activated PEG toward specific amino acids. For the purposes of HBOC development, PEGylation provides a method for optimizing solution colloid osmotic pressure and viscosity, two critical considerations in the preparation of volume-expanding resuscitative fluids. Reactive functional groups are therefore covalently attached to PEG at one or both termini, allowing for surface conjugation, polymerization, or both when reacted with Hb. Bis(maleoylglycylamide) PEG (BMAA-PEG) primarily targets conjugation or polymerization of cysteine residues at the β1/β2 globin chain 93 and 112 positions but can also react avidly with many of Hb’s 44 surfaceaccessible lysine residues, creating a polymer/conjugate mixture consisting of numerous molecular weights (MW) as multiple combinations of modified diaspirin R-R cross-linked Hb (RRDBBF-Hb) are created.5,6 In vitro studies performed in our laboratory revealed that BMAA-PEG covalent addition to RR-DBBF (Poly-RR-DBBF-Hb) increases oxygen affinity and diminishes both cooperativity and pH sensitivity compared to the precursor protein, RR-DBBF-Hb. Moreover, rapid mixing kinetic studies indicated increases in oxygen “off” and “on” rates attributed to Poly-RR-DBBF-Hb compared to RR-DBBF-Hb with Poly-RR-DBBF-Hb exhibiting a slightly greater rate of autoxidation. No differences were observed in the rates of NO binding to RR-DBBF-Hb compared to Poly-RR-DBBF-Hb.7 Rodent studies performed by Abassi et al. evaluated the influence of Poly-RR-DBBF-Hb on systemic hemodynamics, renal function, and aortic inducible NO synthase regulation.8 These data revealed two entirely different hemodynamic responses, Poly-RR-DBBF-Hb caused a significantly lesser hypertensive response following 10% top load (1) Talarico, T.; Guise K. J.; Stacey, C. J. Biochim. Biophys. Acta 2000, 1476 (1), 53-65. (2) Talarico, T.; Swank, A.; Privalle, C. Biochem. Biophys. Res. Commun. 1998, 250 (2), 354-358. (3) Manjula, B. N.; Tsai, A.; Upadhya, R.; Persumalsamy, K.; Smith, P. K.; Malavalli, A.; Vandegriff, K.; Winslow, R. M.; Intaglietta, M.; Prabhakaran, M.; Friedman, J. M.; Acharya, A. S. Bioconjugate Chem. 2003, 14, 464472. (4) Winslow, R. M. Artif. Organs 2004, 28 (9), 800-806. (5) Hai, T. T.; Nelson, D.; Pereira, D.; Srnak, A. Artif. Cells, Blood Substitutes, Immobilization Biotechnol. 1994, 22 (3), 923-931. (6) Hai, T. T.; Nelson, D.; Pereira, D. Bioconjugate Chem. 1998, 9, 645-654. (7) Rogers, M. S.; Brockner Ryan, B.; Cashon, R. E.; Alayash, A. I. Biochim. Biophys. Acta 1995, 1248, 135-142. (8) Abassi, Z.; Kotob, S.; Pieruzzi, F.; Abouassali, M.; Keiser, H. R.; Fratantoni, J. C.; Alayash, A. I. J. Lab. Clin. Med. 1997, 129 (6), 603-610. 10.1021/ac060188q Not subject to U.S. Copyright. Publ. 2006 Am. Chem. Soc.

Published on Web 05/17/2006

compared to RR-DBBF-Hb, while after a 30% exchange transfusion, Poly-RR-DBBF-Hb did not attenuate the hypertensive response caused by RR-DBBF-Hb. These responses appear to be both pharmacological (e.g., NO scavenging) and physicochemical (e.g., oncotic) in nature. Moreover, studies performed by Leppaniemi et al. in rats subjected to vascular injury-induced uncontrolled hemorrhage demonstrated normalized hemodynamics when resuscitated at 5 and 20 mL/kg with a 7 g/dL solution of Poly-RR-DBBF-Hb. Normalization of hemodynamics at these low doses was attributed to the colloid osmotic pressures generated by Poly-RR-DBBF-Hb (52 mmHg at 7 g/dL) restoring mean arterial pressure and cardiac output to baseline.9 Recent investigation in our laboratory used a systematic approach to evaluate modification to HbA0 induced by oxidized raffinose (O-rafffinose)10 and modification to bovine Hb (HbBv) induced by glutaraldehyde, a widely used polymerizing reagent.11 In these experiments, polymeric mixtures were fractionated, functionally characterized, and analyzed by multiple forms of mass spectrometry and amino acid analysis to elucidate the sites of chemical modification. We ultimately determined that O-rafffinose created unexpected non-site-specific modifications to HbA0 in the 2,3-DPG binding pocket and on the surface of HbA0 at βCys93, which affected protein structure, function, and stability that may have lead to the reported toxicity.12 Glutaraldehyde also induced non-site-specific modification in the 2,3-DPG binding pocket and on the surface of HbBv. However, these modifications did not appear to alter protein stability. Thus, using techniques such as mass spectrometry and amino acid analysis affords an ability to elucidate chemical modification to Hb that may be predictive of protein structural and functional abnormalities suggestive of early indicators of toxicity. In the present study, we provide a detailed analysis of another chemical form of modifications to RR-DBBF-Hb imparted by BMAA-PEG. The nature of these modifications is highly nonspecific, yielding a large distribution of MWs, multiple mass additions on the β1/β2 globin chain Cys93 positions, heme pocket conformational changes, and extensive lysine modification. Overall, these chemical modifications to RR-DBBF-Hb lead to differences in the protein structure of the staring HBOC material (RR-DBBF-Hb) and the end product HBOC (Poly-RRDBBF-Hb). Conclusions from these results provide a better understanding of chemical modification and the influence of Poly-RR-DBBF-Hb used in previously reported in vitro and in vivo studies and provides insight and a new interpretation of previous results. EXPERIMENTAL PROCEEDURES Hemoglobin Solutions. Highly purified HbA0 was a kind gift from Hemosol Inc. (Toronto, Canada), and RR-DBBF-Hb and Poly-RR-DBBF-Hb were kind gifts from the Walter Reed Army Institute of Research (Washington D.C.) and Baxter Health Care (Round Lake, IL). The cross-linking reagent, bis(3,5-dibromosalicyl) fumarate was used in the preparation of RR-DBBF-Hb and (9) Leppaniemi, A..; Raluan, S.; Burris, D.; Pikoulis, E.; Ratigan, J.; Waasdorp, C.; Hufnagel, H.; Malcolm, D. Trauma 1996, 40 (2), 242-248. (10) Boykins, R. A.; Buehler, P. W.; Jia, Y.; Venable, R.; Alayash, A. I.. Proteins 2005, 59, 840-855. (11) Buehler, P. B.; Boykins, R. A.; Jia, Y., Norris, S.; Freedberg, D.; Alayash, A. I. Anal. Chem. 2005, 77 (11), 3466-3478. (12) Greenberg, A. G.; Kim, H. W. Crit. Care 2004, 8, S61-S64.

specifically links the two R subunits via Lys99 when reacted with deoxy-Hb. Details of the solution’s purity, preparation, and characterization have been previously outlined 13 and are based on the chemistry described by Chatterjee et al.14 Poly-RR-DBBFHb was prepared by reacting RR-DBBF-Hb with BMAA-PEG as detailed by Hai el al.5,6 Size Exclusion Chromatography. HbA0, RR-DBBF-Hb, and Poly-RR-DBBF-Hb were eluted on a Bio-Sil-TSK-250 (600 mm × 7.5 mm) size exclusion chromatography (SEC) column (Bio-Rad Laboratories Inc., Hercules, CA) attached to a Waters 626 pump and a Waters 2487 dual-wavelength detector and controlled by a Waters 600s controller using Millenium32 software (Waters Corp., Milford, MA). The running buffer consisted of 0.1 M NaH2PO4, pH 6.5, pumped at a rate of 0.5 mL/min, and the absorbance was monitored at 214 and 280 nm. Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis. SDS-PAGE was performed based on the method of Laemmli15 using aliquots of each Hb (10 µg) diluted with Trisglycine SDS buffer. Samples were electrophoresed on a 4-20% polyacrylamide gradient Tris-glycine gel 1.5 h at 125 V using a Novex X Cell SureLock electrophoresis cell. The gel was stained in Coomassie blue overnight after which gels were destained for ∼5 h with deionized water and dried with Novex Gel-drying solution for 15 min. Size Exclusion Chromatography-Multiangle Laser Light Scattering (SEC-MALLS). Molecular size and polydispersity were determined for each Hb fraction by SEC coupled to MALLS. SEC-MALLS data were acquired and analyzed using a Wyatt Technology Corp. Dawn-DSP with Astra version 4.73.04 software. Samples were injected in 100-µL volume and separated on a Shodex OHPAK SB-806HQ column (300 × 8 mm), eluted in 50 mM sodium phosphate, pH 6.9, at 0.5 mL/min. The sample passed through a Dawn DSP laser photometer for MALLS and then an Optilab DSP interferometric refractometer to determine concentration. Both detectors operate at a wavelength of 633 nm. All SECMALLS measurements were made at ambient temperature. Circular Dichroism Spectroscopy. A Jasco J-710 spectropolarimeter was used to compare the CD spectra of HbA0, RRDBBF-Hb, and Poly-RR-DBBF-Hb at equal concentrations (0.1 mM heme) in 0.1 M pH 6.85 Hepes buffer. Measurements were made in the absence and presence of a 5-fold molar excess of the allosteric effector, inositol hexaphosphate (IHP). Acid Hydrolysis and Amino Acid Analysis: 5.7 N HCl Hydrolysis. Approximately 100 µg of each aqueous soluble Hb was placed in glass vials and evaporated to dryness in a Savant vacuum centrifuge. Samples were hydrolyzed in 20 µL of 5.7 N HCl containing 0.2% phenol and flushed with nitrogen prior to heating at 150 °C for 90 min. Following 90 min of heating, glass vials were removed from the heating block, cooled, and evaporated to dryness. The hydrolysate was dissolved in 200 µL of 0.2 N sodium citrate buffer, pH 2.2, syringe filtered, and analyzed with a Hitachi amino acid analyzer (model L-8800). The yields of lysine in HbA0, RR-DBBF-Hb, and Poly-RR-DBBF-Hb were cal(13) Highsmith, F. S.; Driscoll, C. M.; Chung, B. C.; Chavez, M. D.; Macdonald, V. M.; Manning, J. M.; Lippert, L. E.; Berger, R. L.; Hess, J. R. Biologicals 1997, 25, 257-268. (14) Chatterjee, R.; Welty, E. V.; Walder, R. Y.; Pruitt, S. L.; Rogers, P. H.; Amone, A.; Walder, J. A. J. Biol. Chem. 1986, 261, 9929-9937. (15) Laemmli, U.K. Nature 1970, 227, 680-685.

Analytical Chemistry, Vol. 78, No. 13, July 1, 2006

4635

Figure 1. Size exclusion chromatography. Molecular size distribution of (‚‚‚‚) HbA0, (-) RR-DBBF-Hb, and (-‚-) Poly-RR-DBBF-Hb was performed on a Biosep S3000 (600 mm × 4.5 mm) size exclusion column with a running buffer consisting of 0.1 M sodium phosphate, pH 6.5, at a flow rate of 0.5 mL/min. The baseline separation of Poly-RR-DBBF-Hb from HbA0 and RR-DBBF-Hb suggests the absence of low molecular weight (tetrameric) species in the solution of Poly-RR-DBBF-Hb.

culated on the basis of their molar ratios relative to an amino acid standard normalized to alanine.16 The yield of lysine in each Hb sample was compared to that of the theoretical value for HbA0 (44 residues). Each Hb solution was hydrolyzed and analyzed a minimum of four times, and values of each analysis were averaged. Hemoglobin Digestion. Hb samples (1 nM protein) were denatured with 6 M urea in 0.5 M ammonium bicarbonate buffer, reduced with 10 µL of 100 mM dithiothreitol at 50 °C for 60 min, and alkylated with 10 µL of 500 mM iodoacetamide for 30 min at room temperature. Enzymatic digestion was performed by adding trypsin (Modified sequence grade, Roche Diagnostics, Indianapolis, IN) to 0.1 M ammonium bicarbonate, diluted (5-fold), reduced, and alkylated Hb in a 1:1 nM ratio. The reaction was performed at 37 °C over a 7-h period at which time digestion was quenched by adding 10 µL of glacial acetic acid. Samples were lyophilized, reconstituted with either 0.1% Trifluoroacetic acid (TFA) or 1% acetic acid and analyzed by MALDI-TOF and LC-MS, respectively as previously described.17 Regions of protein modification analysis primarily focused on the peptide known to contain the most accessible cysteine residue (βCys93). MALDI-TOF Mass Spectrometry Analysis. Hb solutions were desalted using an analytical reversed-phase C18 HPLC column (4.6 × 250 mm, Vydac, Hesperia, CA) and a gradient generated with 0.1% TFA solution and acetonitrile (AcCN) in 0.1% TFA. AcCN was evaporated from the collected material, and samples were pipetted (1 µL) onto a stainless steel MALDI-MS sample plate and mixed with 1 µL of 3,5-dimethoxy-4-hydroxycinnamic acid (sinapinic acid; Sigma Chemical Co., St. Louis, MO) saturated in 50% AcCN/0.1% TFA acid. The sample/matrix were air-dried and analyzed on a PerSeptive Biosystems DERP MALDITOF mass spectrometer using Data Explorer version 5.0 software (Applied Biosystems, Framingham, MA) operated in the linear mode. Trypsin digests (10 µL of sample) were desalted using C18 ZipTips according to instructions (Millipore Corp., Bedford, MA) followed by evaporation to dryness. The desalted tryptic peptides were redissolved in 0.1% TFA and mixed as above with R-cyano(16) Liu, T-Y.; Boykins, R. A. Anal. Biochem. 1989, 182, 383-387. (17) Schevchenko, A.; Chernushevich, I.; Wilm, M.; Mann, M. Methods Mol. Biol. 2002, 146, 1-16.

4636 Analytical Chemistry, Vol. 78, No. 13, July 1, 2006

4-hydroxycinnamic acid (Sigma Chemical Co.) as the matrix for optimum ionization of peptides of 5000 kDa. LC-ESI Tandem Mass Spectrometry Analysis. Nano LC-MS/MS analysis of Hb peptides was performed using a capillary HPLC coupled directly to a quadrupole ion trap (LCQ Deca) or a quadrupole-TOF (Qstar LCMS/MS) mass spectrometer (ThermoFinnigan, San Jose, CA, or Applied Biosystems, Foster City, CA). Peptide digests were dissolved in 0.01% TFA/0.05% acetic acid prior to column injection. A gradient from 0 to 60% (0.01% TFA/0.05% acetic acid and 0.01% TFA/0.05% acetic acid in AcCN) was used for peptide separation using a C18 column (15 cm × 75 nm, Vydac). The ion trap mass spectrometer was optimized with the nanosource with a spray voltage of 2.5 kV and a heated capillary temperature of 180 °C. Ions were isolated with a mass isolation width (m/z) set at 2.0; AGC mode on; activation Q set at 0.25; activation time, 30 ms; and normalized collision energy set to 35% with a default charge set at +2. Scans were generated in both MS/MS and zoom scan mode of analysis (LCQ Deca). The Qstar mass spectrometer was optimized using the nanosource with positive TOF-MS plus three product ion experiments. Information dependent acquisition (i.e., number of product ion scans) switch criteria was established for ions of >300 mz and for ions of 800 000 (data not shown) indicated the same results. These data, consistent with SDS-PAGE also confirm the concept of non-site-specific modification imparted by reaction of BMAA-PEG with RR-DBBF-Hb as no abundant large m/z [M + H] ions are observed in any of the MALDI-MS spectra, while SEC-MALLS clearly indicated a large polydispersed range of high molar masses (200-4000 kDa) for Poly-RR-DBBF-Hb. LC-ESI Tandem Mass Spectrometry Analysis. Modification to the β1- and β2-globin chain peptides obtained following trypsin digestion and containing the primary target amino acid for BMAA-PEG (βCys 93) was evaluated via LC-MS. The βCys 93 peptide is located in the trypsin-digested fragment containing the amino acid sequence G83TFATLSELHC93DK95 (m/z ) 1478.6, [MH]+). The oxy and deoxy forms of Hb βCys93 and βLys95 (the cleavage site for trypsin) are within ∼10 Å of each other; thus, the covalent addition of BMAA-PEG would likely prevent cleavage at the βLys95. The next downstream sequence of amino acids L96HVDPENFR104 (m/z ) 1126.2, [MH]+) would therefore make up a larger mass conjoined peptide with BMAA-PEG bound to it ((1) G83TFATLSELHC93DK95 - (2) L96HVDPENFR104, (1) 1478.6 + (2) 1126.2 ) 2604.8 mass units). Figure 6 shows the LC-MS total ion current (A, 30-40-min segment) and (B, 40-50-min segment) of peptide G83TFATLSELHC93DK95LHVDPENFR104 (theoretical m/z ) 2604.8) with the addition of several deferring MW BMAA-PEG species (∼3000-4000 kDa). Amino Acid Analysis. Determination of Lysine Content. Amino acid analysis was performed exclusively for lysine content in HbA0, RR-DBBF-Hb, and Poly-RR-DBBF-Hb. As indicated in Figure 5A and B, no mass additions to the R/β chains of HbA0 were observed; nor were unexpected mass additions to the β-globin chains or the cross-linked R-globin chains (R1-R2) of RRDBBF-Hb observed in the MALDI-MS spectra. As a result, it is unlikely that cysteines are modified in either of the two proteins 4640 Analytical Chemistry, Vol. 78, No. 13, July 1, 2006

Figure 7. Amino acid analysis (lysine content). Theoretical value for lysine residues in HbA0 is 44 (‚‚‚‚‚). HbA0 contained 45 unmodified lysine residues as determined by amino acid analysis, while RR-DBBF-Hb was minimally modified (2 lysine residues) as crosslinking occurs internally between lysine 99 positions of the R-globin chains. Lysine modification in Poly-RR-DBBF-Hb was found to be significantly greater (12 lysines) when compared to HbA0 and RR-DBBF-Hb. Statistical differences in lysine content between were determined by ANOVA.

(HbA010 or RR-DBBF-Hb23). Figure 7 shows that all 44 lysine residues are unmodified in both HbA0 with approximately two lysines modified in RR-DBBF-Hb. This is expected as R-R crosslinking occurs between RLys99. However, in Poly-RR-DBBF-Hb, a mean of 12 ( 1 lysine residues (27%) are modified, indicating significant conjugation/polymerization involving BMAA-PEG reaction with lysine rather than cysteine residues. It has been reported that two β-globin cysteine residues (likely βCys93) predominantly react with BMAA-PEG based on determination of thiol content.5 Our data indicate that two cysteine residues are likely modified in a limited number of RR-DBBF-Hb molecules based on MALDI-MS and LC-MS data. Moreover, Table 1 shows the ASAs of lysine and cysteine residues for R- and β-globin chains in deoxy-Hb and, therefore, the potential for reactivity with maleimide-modified PEG (e.g., BMAA-PEG) based on accessibility. It is well established that maleimide will only react with the -amino groups of lysine or the thiol groups of cysteine within Hb. Hemoglobin contains six cysteine residues per tetramer, and in the deoxy conformation, their sulfhydryl groups are generally less accessible to modifying agents. However, the two cysteines located at the β-globin chains 93 positions are particularly accessible and reactive when Hb is in its oxy conformation. The reaction of oxyHb with monofunctional maleimide PEG results in the conjugation of only two copies of PEG 5000 chains per Hb as reported for other forms of surface modifications of human hemoglobin with PEG.24,25 It is important to note that preparation of Poly-RR-DBBF-Hb begins with deoxyRR-DBBF-Hb,5 Thus ASA values are likely somewhat altered based on the fumarate link between R-globin chains; nonetheless, the protein remains in functional T conformation (deoxy-Hb). Our data indicate the predominance of modification imparted by (23) Bossi, A.; Patel, M. J.; Webb, E. J.; Baldwin, M. A.; Jacob, R. J.; Burlingame, A. L.; Righetti, P. G. Electrogelphoresis 1999, 20, 2810-2817. (24) Reiss, J. G. Chem. Rev. 2001, 101, 2797-2919. (25) Veronese, F. M.; Pasut, G. Drug Discovery Today 2005, 10 (21), 1451-1458.

Table 1. Assessable Surface Areas (ASA) of Lysine and Cysteine Residues within Globin Chains globin amino residue chain acid no. R1 R1 R1 R1 R1 R1 R1 R1 R1 R1 R1 R1 R2 R2 R2 R2 R2 R2 R2 R2 R2 R2 R2 R2

Lys Lys Lys Lys Lys Lys Lys Lys Lys Lys Lys Cys Lys Lys Lys Lys Lys Lys Lys Lys Lys Lys Lys Cys

7 11 16 40 56 60 61 90 99 127 139 104 7 11 16 40 56 60 61 90 99 127 139 104

ASA 29.069 35.182 25.869 32.232 64.285 67.444 30.297 64.327 54.154 3.9680 3.3823 0.000 32.968 25.138 55.951 28.399 59.954 32.509 35.170 66.847 47.186 3.9418 3.3827 0.000

globin amino residue chain acid no.

ASA

Lys Lys Lys Lys Lys Lys Lys Lys Lys Lys Lys Cys Cys Lys Lys Lys Lys Lys Lys Lys Lys Lys Lys Lys Cys Cys

63.226 30.101 53.157 51.284 64.252 38.316 50.632 64.416 67.065 14.679 30.029 3.9397 0.8177 52.873 35.200 60.538 58.553 54.624 29.445 58.872 64.157 67.209 17.591 48.821 5.6254 0.9561

β1 β1 β1 β1 β1 β1 β1 β1 β1 β1 β1 β1 β1 β2 β2 β2 β2 β2 β2 β2 β2 β2 β2 β2 β2 β2

8 17 59 61 65 66 82 95 120 132 144 93 112 8 17 59 61 65 66 82 95 120 132 144 93 112

BMAA-PEG is on lysine rather than cysteine and the nature of the overall reaction appears to be very nonsite specific. CONCLUSION The present study evaluated the influence of BMAA-PEG modification on RR-DBBF-Hb with specific focus on MW, solution polydispersity, heme pocket conformation, amino acid modification, and specificity of chemical modification. This experimental approach serves as an example of the necessity to chemically characterize HBOCs in order to gain a full understanding of functional activity and possible toxicity. Experimental data indicate that BMAA-PEG reacted with RR-DBBF-Hb to create Poly-RR-DBBF-Hb results in a HBOC solution with a wide range of MW (200-4000 kDa) leaving essentially no RR-DBBF-Hb (i.e., tetrameric Hb) unreacted. The polydispersity of Poly-RRDBBF-Hb compared to HbA0 and RR-DBBF-Hb was also found to substantially increase, indicating that BMAA-PEG is a nonsite-specific modifying agent. Based on functional data (O2 affinity, O2 off and on rates, and heme iron oxidation) from previous experiments, data suggest a more open and accessible heme pocket in Poly-RR-DBBF-Hb compared to RR-DBBF-Hb and HbA0.7 Our circular dichroism spectrophotometry data indicate (26) Doyle, M. P.; Apostol, I.; Kerwin, B. A. J. Biol. Chem. 1999, 274 (4), 25832591. (27) Rohlfs, R. J.; Bruner, E.; Chiu, A.; Gonzales, A.; Gonzales, M. L.; Madge, M. D.; Vandegriff, K. D.; Winslow, R. M. J. Biol. Chem. 1998, 273 (20), 12128-12134. (28) Acharya, S. A.; Friedman, J. M.; Manjula, B. N.; Intaglietta, M.; Tsai, A. G.; Winslow, R. M.; Malavalli, A.; Vandegriff, K. D.; Smith, P. K. Artif. Cells, Blood Substitutes, Immobilization Biotechnol. 2005, 33 (3), 239-255. (29) Cabrales, P; Tsai, A. G.; Winslow, R. M.; Intaglietta, M. Am. J. Physiol. Heart Circ. Physiol. 2005, 289 (6), H2392-H2400. (30) Wettstein, R.; Tsai, A. G.; Erni, D.; Winslow, R. M.; Intaglietta, M. Crit. Care Med. 2003, 31 (6), 1824-1830.

that the changes brought about by PEGylation of DBBF may have also impacted the heme environment as reflected by the extent of the shift in the Soret CD spectrum of the oxy derivative of PolyRR-DBBF-Hb compared to that RR-DBBF-Hb. Whether these changes reflect an enhanced accessibility of the heme or symmetry difference in the heme environment is unknown at this stage. However, using rapid kinetic techniques, we have previously shown that this form of modification alters ligand binding kinetics of Poly-RR-DBBF-Hb compared to RR-DBBF-Hb, indicating a more open heme pocket environment.7 Extensive characterization of modified protein regions and amino acids in Poly-RRDBBF-Hb using several methods of mass spectrometry and amino acid analysis indicate some modification to the tryptically digested βCys93-containing peptide. However, predominant modification occurred at lysine residues in RR-DBBF-Hb (∼12 of 44). Previous studies suggest that larger MW HBOCs void of tetrameric species are required to blunt or eliminate NO interaction and resulting hypertension.26 This concept is based on the idea that tetrameric Hb can cross the vascular endothelium and scavenge NO at its sites of production. Alternatively, Rohlfs et al. suggest a MW-independent mechanism for HBOC-induced NO scavenging.27 Data presented here support independence of HBOC MW on hypertension when coupled with a direct head-to-head comparison of RR-DBBF-Hb and Poly-RR-DBBF-Hb performed by Abassi et al., indicating no hemodynamic differences between RR-DBBF-Hb and Poly-RR-DBBF-Hb following a 30% exchange transfusion.8 Interestingly, the same group demonstrated an attenuated blood pressure response to Poly-RR-DBBF-Hb versus RR-DBBF-Hb following a 10% blood volume top load.8 Studies performed in a hamster skin-fold model indicate site-specific PEG surface conjugation limits Hb-mediated vasoactivity in toploaded, exchange-transfused, and hemorrhaged animals.28-30 Sitespecific surface conjugation has been shown to increase Hbs radius of hydration, oncotic properties, and viscosity requiring less gram quantities of Hb for infusion.4,28 Non-site-specific PEGylation of RR-DBBF-Hb likely generates solution properties that differ from those generated by site-specific PEGylation of Hb. It is therefore possible that solution properties imparted by sitespecific PEGylation account for favorable microcirulatory events and less systemic hypertension in the hamster skin-fold model. Finally, the present study demonstrates the need for a critical evaluation of chemical alteration of Hb intended for use as HBOCs to aid in understanding functional, biochemical, and physiological properties. ACKNOWLEDGMENT We thank Rick Venable in the Laboratory of Biophysics/ CBER/FDA for performing accessible surface area calculations for cysteine and lysine in R- and β-globin chains of hemoglobin. We also thank the contribution of resources by Dr. Daro´n I. Freedberg in the Laboratory of Biophysics/CBER/FDA. The opinions and assertions contained herein are the scientific views of the authors and are not to be construed as policy of the United States Food and Drug Administration.

Received for review January 27, 2006. Accepted April 18, 2006. AC060188Q Analytical Chemistry, Vol. 78, No. 13, July 1, 2006

4641