How to digest gargantuan data on red cell aging

In this issue of Blood, Nemkov et al used an interdisciplinary approach, incorporating omics analyses, animal models, and genome-wide association studies, to deeply probe the genetic factors that influence the cell metabolism during aging of red blood cells (RBCs).¹ They identified l-carnitine metabolism as the most reproducible pathway involved across multiple blood donations and highlighted a link between this pathway, ex vivo aging of RBCs during storage, in vivo aging of RBCs in circulation, and genetic polymorphisms and their clinical impact after transfusion.

Research on mature, healthy RBCs has provided a plethora of aging markers and mechanisms, especially in the context of transfusion medicine.² During their ex vivo journey in the bag, RBCs leave a dynamic system to stay in a cold acidic plasma-reduced environment until transfusion. Under these conditions, metabolic depletion and alteration, as well as exposure to oxygen, create an imbalance between the antioxidant system and oxidation (namely oxidative stress) that participates in the development of storage lesions.² This cascade of events includes reversible and irreversible modifications in 3 phases: the first transition involving reversible lesions occurs around the second week and the second transition involves irreversible lesions around 4 to 5 weeks of storage.³ Kinetics of these storage lesions is modulated by several genetic and nongenetic factors from the donor and the recipient,^4,5 which results in clinically meaningful changes.⁶ The volume of generated results over the years and especially the gargantuan omics data have been and are still difficult to digest. Will we be able to tease apart the mechanisms of RBC aging, to identify each component of this complex cornucopia?

To take on this daunting task, the research team investigated the molecular phenotype from same donors across multiple donations within the Recipient Epidemiology and Donor Evaluation Study. They carried out a metabolomic analysis of 2 independent donations (index and recall donors, n = 643) and showed that 80% of the measured metabolites were significantly reproducible across donations. Carnitine synthesis and acylcarnitine metabolism were the most reproducible pathways. Consequently, a targeted data analysis of the index cohort (13 091 donors) highlighted a donor subset with high end-of-storage levels of l-carnitine (day 42 postdonation). Interestingly, the l-carnitine level increased with donor age and was higher in male and donors of Asian descent. The recall set confirmed these observations.

Behind these demographics, the most frequent single nucleotide polymorphism (SNP) identified from a genome-wide association study mapped a missense mutation on l-carnitine transporter SLC22A16; this SNP is underrepresented in donors of Asian descent. Correlation analysis between allele frequencies of the 2 top SNPs and omics data revealed a negative association between allele copies and acylcarnitine pools. Of interest, the allele copies were also associated with higher osmotic hemolysis in donors.

A similar genome-wide analysis based on end-of-storage l-carnitine levels in genetically diverse mice showed a significant genome-wide adjusted association with polymorphisms in a region coding for the SLC22A5 transporter. In transfused mice high levels of hydroxyacyl carnitines and short-chain acylcarnitines (breakdown of oxidized fatty acids) were correlated to poor posttransfusion recovery (PTR). The authors also investigated the RBC carnitine pool during in vivo aging. Using sequentially biotinylated RBCs in mice, young and old cells were sorted. l-carnitine and acylcarnitine derivatives were clearly depleted during in vivo aging. A decreased level of l-carnitine was confirmed in human RBCs sorted by density gradients (the densest, the oldest). Moreover, tracing experiments with these sorted cells showed an impairment of the Lands cycle in old RBCs in response to oxidative damage, in agreement with lipid peroxidation during storage. Finally, supplementation of the storage medium with l-carnitine enriched the carnitine pool and improved PTR in transfused mice.

In humans, the observed depletion of carnitine pools correlates with high hemolysis and microvesiculation, known markers of storage lesions. Transfusion of an end-of-storage RBC unit with a low level of carnitine leads to a significantly lower hemoglobin increment in the recipient. Such correlations are consistent with extravascular hemolysis of damaged RBCs at the end of the storage.^7,8 

Taken altogether the presented results support the primary role of carnitine metabolism in preventing oxidative damages and lipid peroxidation (in the absence of mitochondria in mature RBCs),⁹ explaining intravascular and extravascular hemolysis. A higher level of acylcarnitines is associated with lower hemolysis, better storage, and higher hemoglobin increment in recipients. Nevertheless, this mechanism has to be balanced with other observations (eg, sex-related parameters). For instance, l-carnitine is higher in male than in female donors (as is urate, a known antioxidant), whereas hemolysis is known to be lower in women.⁴ This observation does not fully match with carnitine data. Other parameters such as oxygen saturation, lower in female compared with male donors,¹⁰ might be of importance regarding the presence of oxidative damages during ex vivo RBC aging and could explain the lower hemolysis observed in RBC units from female donors.

As for the impact in transfusion medicine and for blood banks, strategies to improve the safety and efficiency of transfusions are multipronged. Obviously, a unique process that guarantees an optimal quality is desirable. However, a targeted preparation of the donated blood that considers the donors’ and receivers’ characteristics and processing (eg, blood irradiation) could be beneficial for storage, as discussed in the present article.¹ Nevertheless, the cost related to blood preparation and logistics are nonnegligible parameters that must be considered. As the management of oxidative damages in RBCs, it is a question of finding the right balance.

Did D’Alessandro and colleagues digest these data? Yes, they succeeded with this in-depth analysis of RBC aging. But I would include the word “partially” since several additional mechanisms have to be included and some of them are still under investigation by this team and other research laboratories. The transfusion community and, in a broader sense, the hematology community are waiting for the next results of these fascinating studies and the biological interpretation of these gargantuan molecular and cellular data.

Conflict-of-interest disclosure: The author declares no competing financial interest.