(160o) Bench-Scale Production of Glutaraldehyde Polymerized Bovine Hemoglobin with Tunable Biophysical Properties As a Red Blood Cell Substitute

Background—Research in the field of hemoglobin-based oxygen carriers (HBOCs) as red blood cell (RBC) substitutes has been persistently overshadowed by safety issues, such as vasoconstriction, systemic hypertension, and oxidative tissue injury, which are primarily explained by the extravasation of low molecular weight (MW) HBOCs into the tissue space. To date, there are no FDA certified RBC substitutes for clinical applications in the United States. The safety issues that have persisted in the development of HBOCs prevents HBOCs from being useful for clinical applications. Previously, our lab investigated the relationship between the MW and vasoactivity of low O₂-affinity, tense-state, glutaraldehyde polymerized bovine hemoglobin (T-PolybHb). We observed that high MW T-PolybHb showed reduced vasoactivity in a top-load model. Hence, our data suggests that the negative vasoactive side effects that have stymied HBOC development can be eliminated by increasing the PolybHb MW. To further optimize this material, we have synthesized a library of high MW T-PolybHb and high O₂-affinity, relaxed-state PolybHb (R-PolybHb) of varying sizes and have characterized their biophysical properties. Furthermore, we have conducted an analysis of the procedural data i.e. meta-data of polymerization to evaluate product yield and quality of our current bench-scale polymerization system.

Methods— Citrate anticoagulated whole blood was collected aseptically from adult cattle (Quad Five, Ryegate, MT). Bovine RBCs were separated and washed by centrifugation with 0.9% saline, and lysed with 3.75 mM phosphate buffer pH 7.4. Hollow fiber (HF) modules with a molecular weight cutoff of 500 kDa and 50 kDa were used to purify and concentrate bovine hemoglobin (bHb) by tangential flow filtration (TFF).

Complete deoxygenation of bHb prior to and during polymerization is required for synthesis of authentic T-PolybHb. To obtain complete deoxygenation of bHb, 29-33 g of bHb at 20.0 mg/ml was placed into an airtight, amber-tinted reactor vessel with continuous stirring, and pumped through a gas exchange membrane swept with N₂. Samples for pO₂ measurements were acquired through a needleless valve, which enabled samples to be taken without introducing oxygen into the syringe. When the pO₂ was < 20.0 mm Hg, 300 mg of sodium dithionite in N₂ purged PBS pH 7.4 was injected into the bHb solution via the needleless valve. This bolus injection of sodium dithionite was allowed to mix for 15-20 minutes, and then, followed by four additional 1 mL injections of 50.0 mg sodium dithionite to ensure complete deoxygenation of the bHb solution. Polymerization of T-PolybHb was not initiated until pO₂ = 0.0 mm Hg and deoxygenation was maintained throughout polymerization by continuous purging of the reactor headspace with N₂. The same reactor vessel configuration was used for complete oxygenation of bHb to prepare for the synthesis of R-PolybHb. Instead, the gas exchange membrane was swept with O₂. Polymerization was not initiated until the pO₂ was > 745 mm Hg and oxygenation was maintained throughout polymerization by continuous purging of the headspace of the reactor vessel with O₂. Glutaraldehyde was used as the crosslinking agent. For both R- and T-PolybHb, bHb was polymerized with glutaraldehyde at glutaraldehyde to bHb molar ratios of 25:1, 30:1, and 35:1. Glutaraldehyde was added dropwise to the bHb solution at a rate of 2 mL/min for 50.0 mL and allowed to react for 2 hours at 37 °C. For quenching the polymerization reaction, a bolus addition of NaCNBH₃ was injected into the reactor vessel to reduce Schiff bases and mixed for 30 minutes as the reactor temperature cooled to room temperature. NaCNBH₃ was prepared at a 7:1 ratio to glutaraldehyde in PBS pH 7.4. The sealed reactor was then placed in a refrigerator at 4^oC overnight.

Both R- and T-PolybHb were sterile filtered via TFF on a 0.2 µm HF module. The PolybHb was then buffer exchanged with a modified lactated Ringer’s buffer pH 7.4 via diafiltration. T-PolybHb 35:1 and 30:1 and R-PolybHb 30:1 and 25:1 were diafiltered and concentrated on a 500 kDa HF module. R-PolybHb 35:1 was diafiltered and concentrated on a 0.2 µm HF module. T-PolybHb 25:1 was diafiltered and concentrated on a 100 kDa HF module. At least 10 diafiltration cycles were required to render the waste permeate void of heme.

Dynamic light scattering was used to measure the hydrodynamic diameter of the PolybHb. To evaluate the size distribution of different species i.e. monomer, dimer and tetramers, we employed size exclusion chromatography (SEC) coupled with high performance liquid chromatography (HPLC). The oxygen affinity and cooperativity coefficient of the PolybHb was measured on a Hemox Analyzer. Meta-data was analyzed using JMP®, Version <Pro 9>, and SAS Institute Inc., Cary, NC, 1989-2019 was used for a statistical analysis.

Results—Biophysical analysis of PolybHb confirm the linear relationship between polymer size and the glutaraldehyde to bHb molar ratio. By keeping the bHb under N₂ or O₂ prior to and during polymerization, the T-PolybHb and R-PolybHb have low and high O₂-affinity, respectively. The T-PolybHb 35:1 and R-PolybHb 30:1, which are our benchmark products, and possess effective diameters 10 times larger than bHb. With this experimental synthesis protocol, the variation in yield for each batch is relatively low, proving the reproducibility of the current protocol. We also observe lower concentrations of methemoglobin when compared with values reported in the literature for other HBOCs.

In the case of the meta-analysis, we find the number of diafiltration cycles is highly correlated with the final product yield, as well as, the concentration of methemoglobin (C_met). Obtaining the lowest C_met without compromising the yield is possible by establishing a certain threshold at which the diafiltration process is terminated.

Conclusions—The biochemical and biophysical properties of T-PolybHb 35:1 and R-PolybHb 30:1 demonstrate their potential to be used as a stopgap O₂ therapeutic. Additionally, analysis of the meta-data shows that finding the threshold that optimizes product yield and minimizes C_met can be used to design a better synthesis protocol. Moreover, the biophysical properties of PolybHb e.g. P₅₀ are tunable to some extent by varying the crosslinking ratio. To our knowledge, this is the first comprehensive evaluation of PolybHb synthesis meta-data. Such analysis allows us to optimize the current synthesis protocol towards industrial scale production.