In this issue of Blood, Reisz et al present a proteomic and metabolomic analysis of red blood cells (RBCs) stored in Additive Solution 3 that focuses on the oxidative changes in the active sites of glyceraldehyde 3-phosphate dehydrogenase (GAPDH). The changes accumulate and become irreversible with time but are sufficiently infrequent that it is unclear that they matter for posttransfusion function.1
In 1916, Rous and Turner2 showed that rabbit RBCs stored for 4 weeks in a solution of citrate and dextrose could be given back to the donor rabbits, raising their hematocrit without raising their reticulocyte count or causing bilirubinuria. The next year, Robertson3 used this solution to build the first human blood bank. Anticoagulant citrate allowed donors to be separated from recipients in space, and nutrient dextrose allowed them to be separated in time. The combination allowed the building of blood banks with dependable inventory and quality control, changing transfusion from a desperate clinical procedure to a routine medical event. Subsequent addition of phosphate, adenine, mannitol, and bicarbonate to RBC storage systems allowed longer storage with more cells circulating 24 hours after reinfusion and reduced hemolysis. Today, RBCs can be stored for 6 weeks with 83% recovery and 0.4% hemolysis.4 Hospitals find recipients for more than 99% of RBCs they procure.
The safety of prolonged RBC storage has been questioned because of associations with excess mortality in retrospective epidemiologic studies and specific transfusion-associated injuries in animals.5 The epidemiologic studies are confounded by a general tendency to transfuse sicker patients. The animal studies that link oxidized lipids in the supernatant plasma of stored RBCs to lung injury and link iron release to inflammation have not been repeatable in human recipients. Several larger human randomized controlled trials have shown no effect of transfusing longer-stored RBCs.6
In an e-Blood article associated with this issue, Reisz and colleagues use proteomics and metabolomics to elucidate RBC storage lesions. They were able to identify and follow the concentrations of normal and oxidized proteins. They focused on the glycolytic enzyme GAPDH, and by using 13C-labeled mass standards of active-site tryptic peptides, they were able to demonstrate at least 10 ways that amino acids in the active sites form initially reversible and later irreversible inactivating modifications. However, the total fraction of the enzyme that was inactivated was small (several percent). If all of the 2,200 proteins in the RBC are similarly damaged in the course of storage, there are more than 22 000 additional protein oxidation-based RBC storage lesions to consider. With the metabolomics, they provide a broader view. The labeled metabolic substrate 13C1,2,3 glucose is metabolized differently in the glycolytic and pentose phosphate pathways. Over the first 2 weeks of storage, a large amount of unlabeled lactate was produced from the breakdown of 2,3-bisphosphoglycerate. The amount of 13C1,2,3 lactate produced in glycolysis decreased during the first 2 weeks of storage but was then stable. The amount of 13C1,2 lactate made in the pentose phosphate shunt increased minimally. The ratio of 13C1,2 lactate to 13C1,2,3 lactate increased slowly and modestly over subsequent weeks of storage. This is consistent with the known inhibition of phosphofructokinase at lower pH as the lactate accumulates. Increasing oxygen saturation of hemoglobin from diffusion of atmospheric oxygen into the storage bag may also drive the greater pentose phosphate activity because less deoxy-β-hemoglobin is available to displace metabolic enzymes from the band III protein cytoplasmic tail. These results do not provide much support for the authors’ hypothesis that damage to GAPDH is limiting glycolysis.
The importance of the article by Reisz and colleagues is the elegant demonstration of new “-omic” technologies brought to bear on the questions of RBC storage. These technologies demonstrate atom-by-atom changes in proteins over the course of weeks of cold storage. The metabolomic findings may be the more important because they suggest that RBCs generally maintain their metabolism as well as decreasing pH allows despite the gradual oxidation of their proteins. This is consistent with a demonstration by Meryman et al7 2 decades ago that RBCs could be stored oxygenated for 9 months at refrigerator temperatures with good recovery if the storage solution was changed regularly to prevent the decline in pH. Metabolic support is the critical function of current RBC storage systems.
Critical for understanding the value of the proteomic and metabolomics data collected by Reisz et al is a context for its utility.8 Merely identifying specific new storage lesions has the potential to lead to expensive new tests for blood donors and a reduced RBC supply. Improving RBC storage generally can reduce the burden of effete RBCs and their toxic breakdown products for everyone, and lengthening safe storage can improve the logistics of RBC availability for units of uncommon phenotype and health care in isolated locations.9
Conflict-of-interest disclosure: J.R.H. is the inventor of Additive Solution 7 and receives patent royalties from the US Army and the University of Maryland.