A Genetic Basis for Donor Variation in Generation of Prostaglandins and Leukotrienes in Stored RBCs Using a Mouse Model

RBC transfusion has clear efficacy in treating various types of anemia. However, in recent decades, there is a renewed focus on the potential for negative clinical sequelae from transfusing stored RBCs. Whether or not older, stored RBC units are associated with adverse outcomes remains controversial. However, there are clear cellular and biochemical data that RBC units accumulate particles and molecules over time with known toxicity when administered to animals and/or humans. Among the most potent of these are prostaglandins and leukotrienes (jointly known as eicosanoids), which are potent mediators of inflammation and vascular pathology. Indeed, arachidonic acid (AA), and 5-, 12-, and 15-hydroxyeicsotetranoic acid (HETE) accumulate during storage of human RBCs and are biologically active in priming neutrophils (Silliman et al., Transfusion 2011 51(12):2549-54). It is well known that there is substantial donor-to-donor variation in how well RBCs store from the standpoint of post-transfusion RBC recovery; however, it is unclear whether the accumulation of eicosanoids varies substantially among donors. If differences exist, then it may be useful to screen RBC units prior to transfusion into patients with illnesses likely to be affected by eicosanoid exposure. Using a well characterized mouse model of RBC storage, and different strains of donor mice, we tested the hypothesis that there are genetic determinants affecting eicosanoid levels in stored RBCs.

RBCs from C57BL/6 (B6) and FVB mice were collected in CPDA-1, filter leukoreduced, and stored under conditions previously shown to model human RBC storage. Samples collected on days 0, 5, 9, and 14 were analyzed by small molecule mass spectrometry. The study was repeated 3 times and combined data were analyzed.

AA accumulated over storage time in both B6 and FVB RBC units to a similar level. In contrast, although essentially no accumulation of eicosanoids was observed in B6 RBC units, substantial time-dependent increases (compared to day of collection) were observed in FVB RBC units for 5-HETE (4-fold) and 15-HETE (12-fold). In addition, a greater than 10-fold increase was observed for prostaglandin E2 in FVB RBC units with no detectable prostaglandin E2 in B6 RBC units [ findings were consistent in all 3 experiments and all differences had p values of <.05]. As oxidized phospholipids are better substrates for phospholipases that generate AA, we hypothesized that oxidative damage would be increased in FVB RBCs. Additional metabolomic analysis demonstrated that B6 RBCs had decreased levels of oxidative damage compared to FVB RBCs, including substantially lower levels of 9,10-epoxystearate, an oxidized membrane lipid. Consumption of glucose and production of lactate were similar in both strains; however, B6 RBCs had increased total glutathione (GSH), oxidized GSH, and cysteine-GSH disulfides. No difference in flux through the pentose-phosphate shunt was observed (including similar NADPH levels); this suggests that the increased GSH in C57BL/6 RBCs was due to de novo synthesis and not to regeneration by GSH reductase. C57BL/6 RBCs also had increased levels of natural anti-oxidants, including ergothioneine and alpha-tocopherol. [All the above differences achieved a p value of <.05].

The current findings take advantage of a mouse model of RBC storage. Given that two inbred, genetically distinct strains demonstrate significant differences in the rate and magnitude of eicosanoid generation, these findings strongly support both the existence of donor variability in eicosanoid accumulation during RBC storage and a genetic basis thereof (in the context of a mouse system). In addition to providing a tractable platform to dissect the genetics of donor-specific RBC storage biology, these findings suggest the hypothesis that similar differences in eicosanoid generation occur with human RBC storage. Subsequent analysis of human specimens will need to be performed to test whether these findings extend to stored human RBC units.

No relevant conflicts of interest to declare.