Page 199 - 2020_08-Haematologica-web

by metabolomic phenotyping plasma from rodents, swine, macaques, and humans,2 hemorrhage in macaques was found to recapitulate human metabolic dysfunction in trauma5 more closely than other animal species. In transfusion medicine, the study of non-human primates facilitated a landmark discovery by Landsteiner and Wiener in the late 1930s, i.e,. identification of the Rhesus blood group – after a factor found in Rhesus monkey blood.6
Red blood cell (RBC) transfusions can be modeled in animal species prior to pursuing costly and complex human clinical trials, or when human studies are deemed unethical (e.g., in acute radiation sickness). Nonetheless, animal modeling requires a rational approach toward species selection; in particular, blood group system diver- sity and variations in RBC physiology and biophysical properties across species/strains provide unique chal- lenges when interpreting animal data in transfusion med- icine. Additional complexity is introduced when refriger- ated storage and blood component preservation are part of the experimental design.7 Refrigerated storage of RBC induces a series of biochemical and morphological modi- fications, collectively denoted “the storage lesion.”8 For example, stored RBC progressively lose their capacity to cope with oxidant stress,9 which is paralleled by their decreased ability to sustain energy metabolism.
Metabolic investigations of human RBC units stored using all currently licensed storage additives10-12 helped to identify the impact of processing strategies (including leukoreduction13 and storage solutions14) on the molecular heterogeneity of stored units. Several factors complicate the study of the potential impact of “age of blood” on clinical outcomes.8 For example, units from some donors may store better than others for genetic, dietary, and/or environmental reasons, and the metabolic age of a RBC unit may differ from its chronological age.15 Findings from the Recipient Epidemiology and Donor evaluation Study (REDS-III) demonstrate donor-dependent heterogeneity in the propensity of RBC to hemolyze in vitro in response to storage duration, oxidative stress, and mechanical/osmotic insults.16 Therefore, improving RBC storage quality through increased understanding of RBC metabolism could enhance RBC quality for all donated units, independently of donor-specific factors, and improve transfusion outcomes overall.
Differences in storage quality become even more com- plex when comparing various animal species and strains; therefore, approximating human RBC function and storage outcomes is critical to pre-clinical, proof-of-concept stud- ies. As such, it is important to use pre-clinical models of novel blood transfusion strategies that reliably approxi- mate clinically relevant scenarios in humans. Although murine and canine models of blood storage and transfusion are available,17–21 macaques are generally perceived as a more relevant pre-clinical model,4 because of their evolu- tionary similarity to humans. For example, hematologic parameters in humans and RM are comparable. The RM hematocrit is 43 ± 2% in males and 41 ± 2% in females with corresponding hemoglobin levels of 13.1 ± 0.9 and 12.5 ± 0.2 g/dL, respectively.22 RM RBC distribution width and disc diameter are 13.0 ± 0.7% and 8 μm, respectively, values similar to those of human RBC.22 Furthermore, the average life span of RM RBC is 98 ± 21 days,23,24 which is similar to that of human RBC (100-120 days), and signifi- cantly longer than that of murine RBC (55-60 days25).
However, little is known about refrigerated storage of RM RBC under standard blood banking conditions. Although prior preliminary studies explored similarities and differ- ences in the proteomes of fresh RBC from mice, humans, and macaques,26 to the best of our knowledge, no previous study compared the metabolome and lipidome of human and RM RBC throughout 42 days of refrigerated storage. Thus, the current data provide a comparative analysis of RBC metabolic pathways and lipidomic changes occurring over time and identify dietary and environmental com- pounds unique to each species.
Methods
Extensive methodological details are provided in the Online Supplementary Methods.
Blood collection, processing and storage
Blood from 5-year old RM (n=20; 10 male/10 female) was col- lected into a syringe, using a 20 G needle, from the femoral vein under ketamine/dexmedotomidine (7 mg/kg/0.2 mg/kg) anes- thesia according to the Food and Drug Administration (FDA) White Oak Animal Care and Use protocol 2018-31. All blood donor animals originated from the same colony located on Morgan Island, South Carolina and were naïve to experimenta- tion at the time of blood collection.
The blood from 30- to 75-year old human volunteers (n=21; 11 male/10 female) was collected into a syringe, using a 16 G needle, from the median cubital vein under informed consent according to National Institutes of Health (NIH) study Institutional Research Board #99-CC-0168 “Collection and Distribution of Blood Components from Healthy Donors for In Vitro Research Use” under an NIH-FDA material transfer agree- ment. Blood was collected into acid citrate dextrose, leukofil- tered, and stored in AS-3 in pediatric-sized bags, designed to hold 20 mL volumes, which mimicked the composition of stan- dard full-sized units (i.e., incorporating polyvinylchloride and phthalate plasticizers).
The RBC were stored at 4-6oC for 42 days. The RBC and supernatants were separated via centrifugation upon sterile sam- pling of each unit on days 0, 7, 14, 21, 28, 35, and 42.
haematologica | 2020; 105(8)
Metabolism of stored human and macaque RBC
Ultra-high pressure liquid chromatography - mass spectrometry metabolomics and lipidomics
Frozen RBC aliquots of 50 mL volume were extracted 1:10 in ice-cold extraction solution (methanol:acetonitrile:water 5:3:2).27 Samples were vortexed and insoluble material pelleted, as described elsewhere.28 Analyses were performed using a Vanquish UHPLC coupled online to a Q Exactive mass spec- trometer (Thermo Fisher, Bremen, Germany). Samples were analyzed using a 3 min isocratic condition29 or a 5, 9, and 17 min gradient, as described previously.30,31 Additional analyses, includ- ing untargeted analyses and fragment ion search (FISh) score cal- culation via mass spectrometry,2 were performed with Compound Discoverer 2.0 and LipidSearch (Thermo Fisher, Bremen, Germany). For targeted quantitative experiments, extraction solutions were supplemented with stable isotope- labeled standards, and endogenous metabolite concentrations were quantified against the areas calculated for heavy isotopo- logues for each internal standard.30,31 Graphs and statistical analy- ses (either a t-test or repeated measures analysis of variance) were prepared with GraphPad Prism 5.0 (GraphPad Software, Inc, La Jolla, CA, USA), GENE E (Broad Institute, Cambridge, MA, USA), and MetaboAnalyst 4.0.32
2175