Impact of blood transfusion on gene expression in human reticulocytes

To the Editor: Blood transfusion is a frequently performed therapeutic procedure that requires regular evaluation, particularly for its indications, effectiveness, and risks. There are indeed an increasing percentage of blood transfusions considered to be inappropriate and their efficiency has raised questions. The identification of new specific biomarkers of blood transfusion would be of particular relevance for the monitoring of this method. These markers would also be applicable for anti-doping purposes. Infusion of blood results in a rapid increase of circulating red blood cells (RBCs), which impairs endogenous production and release of immature RBCs. This impact of transfusion is mediated by the suppression of erythropoietin (EPO) 1, 2. We hypothesized that blood re-infusion may cause a decrease in the expression of genes related to structural and functional components of reticulocytes and red blood cells (RBCs). The present study aimed to investigate the transcriptional response of a subset of genes, whose functions are related to reticulocyte metabolism, after autologous blood transfusion (ABT) using the digital multiplex mRNA profiling. Seven healthy male volunteers (age range, 20–35 years; body mass index, 18–30), that were eligible for blood donation according to national regulations were included in the study. Details regarding the clinical trial (NCT02423135) were previously described 2. Briefly, during the control phase, all volunteers were infused with saline solution. Fourteen days later, all volunteers donated one full bag of blood (approximately 500 mL). The concentrated RBCs were stored at ∼4°C until re-infusion 36 days later. RNA expression was measured using Nanostring® nCounter Analysis System (Nanostring Technologies, Seattle, WA). Gene expression analysis was performed at baseline (D-4 and D-1) and after the infusion of 0.5 L of saline solution (0.9% NaCl, BBraun, Cressier, Switzerland) on day 3, 6, and 9 and after re-infusion of one's own blood (0.28 L) on day 6, 9, and 15. Blood samples were drawn into Tempus Blood RNA tubes (Life Technologies, Carlsbad, CA) to stabilize genomic material and were stored at −20°C until further extraction. A one-way ANOVA followed by post hoc pairwise comparisons (t-tests adjusted by Bonferroni corrections and Tukey's Honestly Significant Difference) were used to test differences between samples taken during saline or transfusion phase. Based on the blood transcriptional signature of recombinant human erythropoietin 3, a subset of 45 genes (following removal of the ten housekeeping genes) was selected for evaluation of their expression following ABT. Agglomerative clustering (heat maps) resulted in the identification 27 genes that were commonly down-regulated 6, 9, and 15 days after blood re-infusion in seven volunteers (Supporting Information S1 Table and Supporting Information S1 Fig.). Delta-aminolevulinate synthase 2 (ALAS2), carbonic anhydrase (CA1), and solute carrier family 4 member 1 (SLC4A1) were observed to have the greatest fold-change following blood re-infusion than at baseline. A marked decrease in gene expression was observed 6 days after ABT, although the number of transcripts for each gene was significantly decreased 9 days, after re-infusion of blood (Fig. 1A). At day 15, gene expression remained low, and was not significant. During the control phase, the number of transcripts of the candidate genes did not vary significantly 3, 6, and 9 days and stayed steady after saline infusion (Fig. 1B). Finally, the expression of the housekeeping genes (ACTB, ACTR10, MRFAP1, TBP, TRAP1) remained constant throughout the phases, indicating that the changes observed in gene expression are not cell count-based (Supporting Information S2 Fig.). It suggests that the variations of gene expression observed during the transfusion phase are specific to the re-infusion of blood. A: Transfusion phase. Expression of ALAS2, CA1, and SLC4A1 before (D-4 and D-1) and 6, 9, and 15 days (n = 7) after autologous blood transfusion. The dashed line indicates transfusion. Gene expression fold-changes were calculated on original/normalized mean data and are annotated below the time points 6, 9, and 15 days post-transfusion relative to baseline (mean of the data from 1 and 4 days pre-transfusion). Raw counts were normalized to internal levels of five reference genes, ACTR10, ACTB, MRFAP1, TBP, and TRAP1. The Y axis represents log2-transformed normalized counts of the genes. #P ≤ 0.05, statistically significant difference compared with baseline values. B: Saline phase. Expression of ALAS2, CA1, and SLC4A1 before (D-4 and D-1) and 3, 6, and 9 days (n = 7) after saline infusion. The dashed line indicates saline infusion. Raw counts were normalized to internal levels of five reference genes, ACTR10, ACTB, MRFAP1, TBP, and TRAP1. The Y axis represents log2-transformed normalized counts of the genes. No statistically significant difference was observed between time points. Although they have shed their nucleus, circulating blood reticulocytes still retain quantities of functional residual acid material which is essential for their maturation into erythrocyte 4. These remaining copies contained in circulating reticulocytes are hypothesized to reflect gene expression activity of erythroblast into bone marrow 5. Erythroid precursors express at their surface a receptor specific to EPO. When secreted upon hypoxia, EPO targets developing erythroblasts and controls their differentiation and proliferation. It also contributes to the release of reticulocytes through a diminution of the normal marrow-peripheral blood barrier. Previously, Durussel et al. demonstrated that EPO injection influenced the gene expression in human reticuloctytes 3. As blood transfusion suppresses erythropoiesis through the decrease of EPO concentration 1, 2, similar changes in mRNA expression specific to reticulocytes were expected. Our study demonstrates that the transfusion of autologous blood triggers a down-regulation of genes that are involved in biological processes related to reticulocytes and RBCs. ALAS2 is involved in the heme synthesis, whereas CA1 and SLC4A1 are responsible for the transport of oxygen and carbon dioxide. Thus, these candidate genes are specific to erythropoiesis. The magnitude of the changes of the genes transcripts was more important compared to the small physiological effect of ABT on peripheral blood markers 2. In anti-doping field, autologous blood transfusion is assessed by measuring hematological parameters via the Athlete Biological Passport (ABP) which involves brittle biological materials. It requires costly investments in the pre-analytical steps to ensure the validity of the analyses 6. To overcome actual challenges, our study proposed the inclusion of transcriptomic biomarkers, whose sensitivity is greater than that of classical variables, into the adaptive model of the ABP, coupled with easy-to-use collection blood tubes that stabilize genomic material for up to 5 days at room temperature and for years when kept frozen. However, before a potential integration of these three innovative biomarkers into the adaptive model of the ABP, intrinsic and extrinsic factors that may affect the expression of these genes must also be fully characterized. In summary, our results demonstrate that autologous blood transfusion triggered a down-regulation of genes whose function is linked to the metabolism of immature RBCs. Following autologous infusion of stored RBCs, expression of the genes ALAS2, CA1, and SLC4A1 was markedly decreased. The profiling of reticulocytes transcriptome offers a new clinical way for the study of erythroid biology in response to blood transfusion. Futhermore, these transcriptomic biomarkers may also serve for the detection of autologous blood transfusion in an anti-doping context as they appeared to be more sensitive than classic hematological biomarkers. The authors would like to thank the Genomic Technologies Facility at the University of Lausanne for Nanostring analyses and their assistance for the understanding of nSolver software. Olivier Salamin,1 Laura Barras,1 Neil Robinson,1 Norbert Baume,1 Jean-daniel Tissot,2 Yannis Pitsiladis,3 Martial Saugy,1 And Nicolas Leuenberger1* 1Swiss Laboratory for Doping Analyses, University Center of Legal Medicine, Lausanne and Geneva, Centre Hospitalier Universitaire Vaudois and University of Lausanne, Lausanne, Switzerland; 2Site D'Epalinges, Transfusion Interrégionale CRS, Epalinges, Switzerland; 3FIMS Reference Collaborating Centre of Sports Medicine for Anti-Doping Research, University of Brighton, Eastbourne, United Kingdom Additional Supporting Information may be found in the online version of this article. Contract grant sponsor: World Anti-Doping Agency (WADA); Contract grant number: 12C14NL; Contract grant sponsor: Département Universitaire de Médecine et Santé Communautaire (DUMSC); Contract grant number: 06/2015. Additional Supporting Information may be found in the online version of this article. Supporting Information Figure S1 Supporting Information Figure S2A Supporting Information Figure S2B Supporting Information Table S1 Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.