Cross-linked human hemoglobins have been evaluated for clinical use as circulating oxygen carriers. However, their induction of vasoactivity was sufficiently problematic to lead to the cessation of clinical trials. The source of vasoactivity is likely to be endothelial extravasation causing the scavenging of endogenous nitric oxide. It was recently shown that species that consist of two coupled hemoglobin tetramers do not evoke vasoactivity in a sensitive murine model. Presumably these materials are too large to extravasate. In order to make this class of material more readily available, there is a need for improved methods that can form a cross-linked bis-tetramer without producing smaller species at the same time. A potentially efficient route to cross-linking and coupling two Hb tetramers is through phase-directed copper-catalyzed azide alkyne cycloaddition (PDCuAAC). However, introduction of the necessary azide-containing cross-link gives mixtures of tetrameric and bis-tetrameric proteins, as the PDCuAAC process appears to be limited to only those proteins where the cross-link containing the azide is exclusively within the β-subunits. In order to block formation of the azide cross-link within the α-subunits, subunit-specific introduction of the azide is necessary. This is achieved by blocking reaction at the reactive amino groups of the β-subunits in the site that binds the allosteric activator 2,3-diphosphoglycerate (DPG) with inositol hexaphosphate (IHP), permitting α-selective acetylation with acetyl 3,5-dibromosalicylate. After removal of IHP, reaction with an anionic cross-linker containing an azide group occurs within the β-subunits. The resulting α-acetylated β-β'-cross-linked hemoglobin azide (acHb>-N3) undergoes efficient PDCuAAC with bis-alkynes to produce cross-linked bis-tetramers. Analysis of circular dichroism spectra of the modified species shows that there is little change in the structure of the globin chains as a result of the chemical modifications. The oxygenation properties are consistent with those needed for effective oxygenation in circulation, while the bis-tetrameric structure is sufficiently large to avoid extravasation and depletion of nitric oxide.

Cold static preservation on ice (~4°C) remains the clinical standard of donor organ preservation. However, mitochondrial injury develops during prolonged storage, which limits the extent of time that organs can maintain viability. We explored the feasibility of prolonged donor lung storage at 10°C using a large animal model and investigated mechanisms related to mitochondrial protection. Functional assessments performed during ex vivo lung perfusion demonstrated that porcine lungs stored for 36 hours at 10°C had lower airway pressures, higher lung compliances, and better oxygenation capabilities, indicative of better pulmonary physiology, as compared to lungs stored conventionally at 4°C. Mitochondrial protective metabolites including itaconate, glutamine, and N-acetylglutamine were present in greater intensities in lungs stored at 10°C than at 4°C. Analysis of mitochondrial injury markers further confirmed that 10°C storage resulted in greater protection of mitochondrial health. We applied this strategy clinically to prolong preservation of human donor lungs beyond the currently accepted clinical preservation limit of about 6 to 8 hours. Five patients received donor lung transplants after a median preservation time of 10.4 hours (9.92 to 14.8 hours) for the first implanted lung and 12.1 hours (10.9 to 16.5 hours) for the second. All have survived the first 30 days after transplantation. There was no grade 3 primary graft dysfunction at 72 hours after transplantation, and median post-transplant mechanical ventilation time was 1.73 days (0.24 to 6.71 days). Preservation at 10°C could become the standard of care for prolonged pulmonary preservation, providing benefits to both patients and health care teams.

Donor organ allocation is dependent on ABO matching, restricting the opportunity for some patients to receive a life-saving transplant. The enzymes FpGalNAc deacetylase and FpGalactosaminidase, used in combination, have been described to effectively convert group A (ABO-A) red blood cells (RBCs) to group O (ABO-O). Here, we study the safety and preclinical efficacy of using these enzymes to remove A antigen (A-Ag) from human donor lungs using ex vivo lung perfusion (EVLP). First, the ability of these enzymes to remove A-Ag in organ perfusate solutions was examined on five human ABO-A1 RBC samples and three human aortae after static incubation. The enzymes removed greater than 99 and 90% A-Ag from RBCs and aortae, respectively, at concentrations as low as 1 μg/ml. Eight ABO-A1 human lungs were then treated by EVLP. Baseline analyses of A-Ag in lungs revealed expression predominantly in the endothelial and epithelial cells. EVLP of lungs with enzyme-containing perfusate removed over 97% of endothelial A-Ag within 4 hours. No treatment-related acute lung toxicity was observed. An ABO-incompatible transplant was then simulated with an ex vivo model of antibody-mediated rejection using ABO-O plasma as the surrogate for the recipient circulation using three donor lungs. The treatment of donor lungs minimized antibody binding, complement deposition, and antibody-mediated injury as compared with control lungs. These results show that depletion of donor lung A-Ag can be achieved with EVLP treatment. This strategy has the potential to expand ABO-incompatible lung transplantation and lead to improvements in fairness of organ allocation.