Abstract
Erythropoiesis has been extensively studied using in vitro and in vivo animal models. Despite this, there is still limited data about the gene expression profiles (GEP) of primary (ex vivo) normal human bone marrow (BM) erythroid maturation. We investigated the GEP of nucleated red blood cell (NRBC) precursors during normal human BM erythropoiesis. Three maturation-associated populations of NRBC were identified and purified from (fresh) normal human BM by flow cytometry and the GEP of each purified cell population directly analyzed using DNA-oligonucleotide microarrays. Overall, 6569 genes (19% of the genes investigated) were expressed in ≥1 stage of BM erythropoiesis at stable (e.g., genes involved in DNA process, cell signaling, protein organization and hemoglobin production) or variable amounts (e.g., genes related to cell differentiation, apoptosis, metabolism), the latter showing a tendency to either decrease from stage 1 to 3 (genes associated with regulation of erythroid differentiation and survival, e.g., SPI1, STAT5A) or increase from stage 2 to stage 3 (genes associated with autophagy, erythroid functions such as heme production, e.g., ALAS1, ALAS2), iron metabolism (e.g., ISCA1, SLC11A2), protection from oxidative stress (e.g., UCP2, PARK7), and NRBC enucleation (e.g., ID2, RB1). Interestingly, genes involved in apoptosis (e.g., CASP8, P2RX1) and immune response (e.g., FOXO3, TRAF6) were also upregulated in the last stage (stage 3) of maturation of NRBC precursors. Our results confirm and extend on previous observations and providing a frame of reference for better understanding the critical steps of human erythroid maturation and its potential alteration in patients with different clonal and non-clonal erythropoietic disorders.
Similar content being viewed by others
Introduction
Erythrocytes are currently recognized as bone marrow (BM)-derived blood cells specialized in producing hemoglobin for the transportation of oxygen throughout the body, via peripheral blood (PB). After birth, human erythropoiesis occurs in specialized hematopoietic stem cell (HSC) niches enriched in mesenchymal cells and other stromal cells in the BM1. Following commitment into the erythroid lineage, progressive expansion of the early erythroid committed progenitors occurs, leading to the generation of burst-forming unit-erythroid progenitors (BFU-E) and colony-forming unit-erythroid precursors (CFU-E)2. Subsequently, erythroid precursors sequentially differentiate into morphologically recognizable (nucleated) erythroblasts, and both non-nucleated reticulocytes and erythrocytes, which are released into PB2. During maturation of erythroblasts, progressive decrease in cell size with concomitant nuclear condensation and pyknosis, associated with a decrease in the RNA contents occurs, prior to final extrusion of the remaining nucleus3,4 required to generate mature (non-nucleated) reticulocytes and erythrocytes2.
At present, it is well-known that maturation of early hematopoietic precursors into fully mature erythrocytes involves complex signaling networks (associated with, e.g., cell–cell interactions and cytokine and hormone signaling), that lead to intracellular production of hemoglobin and other erythroid cell components5. Early studies on erythroid differentiation have elucidated the composition of the erythroid cell membrane and the key elements of the red cell metabolism6. Subsequently, several other in vitro and in vivo animal model studies have delineated the different steps of erythroid maturation both in physiological and in pathological conditions7. More recent studies have investigated and identified specific genes, cell signaling pathways and proteins that play a critical role during erythropoiesis, particularly those that are (directly or indirectly) related to the production of heme and hemoglobin8,9. Despite all the above, there is still limited data about the changes on the ex vivo gene expression profiles (GEP) of normal human BM erythroid precursors at distinct stages of maturation, and their correlation with the in vitro and in vivo animal model findings10.
In the present study, we investigated the GEP of purified nucleated erythroid precursors at three different stages of maturation, directly derived from primary (fresh) normal human BM, in the absence of further in vitro culture. Our results provide a frame of reference for better understanding the critical steps of in vivo erythroid maturation, and the potential identification of aberrant GEP in patients with different (clonal and non-clonal) red cell disorders.
Results
Identification of erythroid precursors and their different maturation stages in normal human BM
Overall, three different stages of maturation of NRBC were defined and identified in all BM samples analyzed (Fig. 1), and nucleated cells included from each of such maturation stages, were subsequently FACS-purified. Thus, the earliest stage of erythroid maturation corresponded to BM CD34+ precursors that showed progressively greater expression levels of CD71 and CD105 (stage 1). Subsequently, (stage 2) erythroid precursors showed strong CD71 and co-expression of CD105 and CD36, while they had already downregulated CD34, CD45, CD117, and HLADR expression to undetectable levels (Fig. 1). Further, maturation of erythroid precursors (toward stage 3) was associated with downregulation of CD105, while strong positivity for both CD71 and CD36 was retained (Fig. 1). NRBC precursors from all three stages of maturation were systematically negative for CD33 (Fig. 1).
Protein expression levels for the above listed phenotypic markers (as assessed by flow cytometry), was compared with mRNA expression levels for the corresponding genes, showing highly consistent results (Fig. 2), and, as could be expected, intracellular decrease in mRNA levels frequently preceded that observed in the amount of cell surface membrane protein, as illustrated for, e.g., CD117 in Fig. 2. Thus, CD33, CD34, CD45, and HLADR expression levels were virtually not detected at the mRNA level for any of the three maturation-associated populations of NRBC analyzed, while expression of the CD34, CD45, and HLADR proteins was restricted to the earliest stage of maturation of NRBC precursors, and that of CD33 was systematically absent. In contrast, CD71 and CD36 showed parallel and progressively greater amounts of both mRNA and protein along the erythroid maturation.
Global transcriptional profile of normal human BM NRBC precursors
From all 33,927 genes analyzed, 6569 genes (19%) were expressed in ≥ 1 of the three populations of NRBC analyzed. Almost half of the expressed genes (n = 2928 genes) were expressed across all three maturation-associated subpopulations of NRBC precursors evaluated; in contrast, 148, 85, and 1529 genes showed an expression profile restricted to the early (stage 1), intermediate (stage 2), or late stages (stage 3) of maturation of NRBC, respectively. Another 227, 279, and 307 genes were expressed in stage 1 plus stage 2, stage 2 plus stage 3, and stage 1 plus stage 3 NRBC, respectively (Fig. 3a). According to their biological role, those 6569 genes found to be expressed in NRBC could be classified into nine distinct functional gene groups, that included genes involved in: (i) structural cell organization, (ii) metabolic processes, (iii) DNA/RNA processing, (iv) intracellular transport, (v) cell signaling and protein organization; (vi) cycle cell, (vii) cell differentiation, (viii) apoptosis, and (ix) the immune response.
As expected, a high number of genes expressed by BM NRBC precursors were functionally related to DNA/RNA processing (e.g., HIST1H1E, HIST1H1B, and HIST1H2BM histones) and cell signaling and protein organization (e.g., the RPS20, RPL19, and RPL30 ribosomal protein genes), and they were expressed across all maturation stages of NRBC precursors, although the number of expressed genes within both functional groups slightly increased from stage 2 to stage 3 NRBC precursors (Fig. 3). A GEP similar to that of this later gene group was (i.e., stable GEP during the first two stages of maturation, followed by increased expression in stage 3 NRBC precursors) also found throughout the whole human BM erythroid maturation, but for a lower number of genes, for genes related to (i) cell differentiation (e.g., RPS6, MED1, HCLS1), (ii) metabolic processes, (iii) structural organization of the cell (e.g., EPB42, PLEK2, PDLIM7, ROCK1), (iv) intracellular transport, and (v) apoptosis (e.g., P2RX1, DDIT3, RBM38, CUL4A). In contrast, the number of cell cycle-associated genes expressed at the RNA level already increased in the transition from stage 1 to stage 2 (e.g., CDC20, SIPA1, DTL) human BM erythroid precursors, whereas histone-binding transcriptional activators showed either stable (e.g., KAT2A) or increased (e.g., CYBRD1) mRNA levels at the last stage of maturation of NRBC precursors. Genes involved in the regulation of erythroid differentiation (e.g., the CDK6 and STAT5A genes) and anti-apoptotic (i.e., survival) mechanisms (e.g., DDX42, MEX3D, and SERBP1 genes) were mostly expressed among stage 1 NRBC, while genes involved in the immune response were expressed at relatively low numbers, predominantly in stage 3 precursors (Fig. 3).
GEP of erythroid lineage-associated markers during normal human BM erythropoiesis
Overall, maturation of NRBC in human BM was associated with modulation of erythroid differentiation-associated GEP. Thus, transcriptional factors involved in erythroid specification of hematopoietic stem cells and erythroid differentiation such as the GATA1, STAT2, FOG1, and KLF1 genes, were expressed across all maturation stages, in the absence of PU.1 expression (multipotentiality transcription factor), while expression of the STAT5A gene involved in EpoR signaling progressively decreased with maturation, being absent in stage 3 NRBC (Table 1). Similarly, expression of the GFI1B and TAL1 genes increased in stage 2 NRBC, and either remained stable (GFI1B gene) or increased further (TAL1 gene) thereafter (Table 1). In contrast, JAK2 and PTPN6 were upregulated only in stage 3 NRBC (Table 1). In turn, genes involved in the synthesis of heme such as ALAS2 and ALAS1 reached their maximum levels of expression at the more mature (stage 3) NRBC precursors, whereas expression of enzymes involved in degradation of heme (e.g., the HO1 and HO2 genes) was absent or very low across all three erythroid maturation stages analyzed (Table 1).
Similarly, several genes associated with the homeostasis of iron (e.g., FBXL5, FTL) were predominantly, or exclusively, expressed at the latest stage of maturation of NRBC (stage 3) here analyzed; in contrast, expression of other genes related to iron homeostasis/mitochondrial transportation (e.g., SLC25A38, SCL25A39, and SFXN1), was stable along all erythroid maturation stages. A similarly stable GEP was observed for most genes associated with the oxidative stress response (e.g., UPC2, PRDX3, TXNIP, DUSP1, GAB1, and GPX1), with only a few genes showing expression restricted to stage 3 NRBC (e.g., PARK7, EGLN1, and SOD2) (Table 1). MMP8 and MMP9 were also expressed in all maturation stages of NRBC, but their levels were gradually upregulated in stage 2 and stage 3 NRBC. Finally, hemoglobin genes (e.g., alpha, beta, delta, mu, gamma, and theta1) were already expressed from the earliest stages of maturation (stage 1), across all erythroid populations analyzed (Table 1 and Supplementary Figure 3).
Expression profile of genes related to apoptosis, autophagy, and enucleation of NRBC precursors
Concerning apoptosis-associated genes, PML and CASP-2 were both expressed across the three NRBC stages of maturation analyzed, whereas CASP-3 and CASP-9 were absent and CASP8 expression was restricted to the last maturation stage (Table 1). In turn, genes involved in mitochondrial and ribosomal clearance such as the ATG2, ATG4, ATG9, mTOR, ATG1 (ULK1) and RB1CC1 genes, were progressively upregulated during erythroid maturation, and genes associated with NRBC enucleation (e.g., MED1, MAEA, ID2, RB1, SP1, and SP3) showed a predominant or restricted expression in stage 3 NRBC (Table 1). Interestingly, the MAEA gene—macrophage erythroblast attach gene—coding for a protein that mediates the linkage between erythroblasts and macrophages in the enucleation process during terminal erythroid maturation, displayed a stable expression across all NRBC maturation stages (Table 1).
GEP for immune response-associated markers during normal human BM erythropoiesis
Interestingly, several genes associated with the immune response (e.g., FOXO3, BPI, CAECAM8, S100A8, S100A9, VNN1, and THBS1 genes) were upregulated at the last stage of maturation of NRBC here analyzed. Genes involved in response to proinflammatory cytokines and in activation of the IkappaB kinase (IKK) within the NF-kappaB pathway (e.g., TRAF6 gene), showed restricted expression to more mature (stage 3) NRBC. In turn, genes of the STAT family, such as STAT5B, STAT1, and STAT3, were expressed across all maturation stages, although the later two genes showed slightly decreased levels in stage 2 vs. both stage 1 and 3 NRBC, and the GDF15 gene was absent in all NRBC maturation stages analyzed (Table 1).
Discussion
Up till now, only a few studies have investigated the GEP of normal human BM erythroid precursors10,11,12. However, in these such studies, analysis of the GEP of in vitro differentiated erythroid precursor cells was performed13, which might not directly reflect the real in vivo GEP profiles. Here, we evaluated the ex vivo GEP of distinct maturation-associated subsets of erythroid precursors directly isolated from (fresh) normal human BM. Based on a previously established 8-color antibody combination validated by the EuroFlow consortium14 for the identification of well-defined stages of maturation of NRBC in human BM, three distinct maturation-associated subsets of NRBC were identified for their further isolation and subsequent analysis of their GEP. As previously demonstrated, the three purified NRBC populations corresponded to NRBC precursors presenting morphologic features of proerythroblasts, basophilic, and both polychromatic/orthochromatic erythroblasts, respectively (data not shown)15. Of note, the isolated NRBC populations showed a high purity as confirmed by both flow cytometry and the absence of expression of genes related to other hematopoietic cell lineages, together with strong expression of erythroid-specific and erythroid-associated genes.
Overall, our results suggest that from the less differentiated to the more mature erythroid precursors, less requirements for molecules involved in cell division (e.g., absence/low SMARCB1, PFDN1, TRNP1 expression) exist, in parallel to preserved expression of genes related to cell cycle arrest (e.g., RBM38, CUL4A, KAT2B), and an increased expression of genes involved in both cell cycle regulation (e.g., CDC20, SIPA1, DTL) and cell differentiation (e.g., RPS6, MED1, HCLS1), structural organization (e.g., PLEK2, PDLIM7, ROCK1), and signal transduction16 (e.g., P2RX1); all these later genes were expressed in parallel to genes involved in apoptosis and the immune response as discussed below.
At present, it is well-established that Epo signaling via the EpoR/Jak2/Stat5 pathway that coordinates differentiation, proliferation, and survival of erythroid precursors17,18, is critical for the production of erythroid cells19. Overall, our results showed EPOR gene expression throughout the whole human BM erythroid maturation, with greater levels at intermediate and late stages of differentiation; these results are in contrast with murine studies that show strongly decreased Epo levels at the more advanced stages of erythropoiesis20. Previous studies have shown that the expression of the above genes is strongly modulated by oxygen availability19. Thus, hypoxia alters progression through erythropoiesis, via modulation of the EPOR signaling pathway and induction of specific erythroid transcription factors involved in signaling for the production and accumulation of hemoglobin during erythropoiesis21, such as the GATA1, FOG1, TAL1, KLF1, and GFI1B genes22,23. In this regard, GATA-1 is currently considered to be the major transcriptional factor for erythroid and megakaryocytic differentiation, via repression of genes involved in progenitor multipotentiality like PU.12,16. Besides, GATA-1 also promotes G1 cell cycle arrest during terminal erythroid differentiation via repression of mitogenic genes24,25 and upregulation of anti-apoptotic genes26 and other transcriptional factors (e.g., FOG-1 and KLF-1) involved in the biosynthesis of globins and heme27, which are critical for an efficient production of erythroid cells. In line with this, here we found strong and stable expression across all NRBC maturation stages of the GATA-1 as well as the FOG-1, STAT2, STAT5B, and KLF1 transcription factors, similar to what has been previously described in in vitro human models28. However, despite similarly stable GEP across the whole erythropoiesis were identified for a high number of genes, an overall tendency toward downregulation of multiple genes between stages 1 and 2, together with upregulation of other genes in stage 3 vs. stage 2 NRBC precursors, was observed.
In this regard, as could be expected, genes that code for enzymatic complexes involved in the synthesis of heme such as the ALAS1 and ALAS2 genes29 were highly expressed across all erytropoietic maturation stages analyzed. Thus, ALAS2 expression progressively increased during erythroid maturation, while increase in the expression of ALAS1 was less pronounced, delayed and restricted to the more mature (stage 3) erythroid precursors. Similarly, genes involved in iron assembly, transport and homeostasis (e.g., ISCA1 and ISCA2), as well as in iron transfer and transport (e.g., SLC11A2, SLC25A39, and SLC25A38) progressively increased during erythroid maturation. In turn, proteases involved in extracellular matrix digestion, such as MMP8 and MMP9, were also found to be progressively upregulated in stage 2 and 3 NRBC precursors. MMP9 expression is known to be suppressed in the presence of heme oxygenase-1 (HO-1)30, an enzyme involved in the catabolism of heme with antioxidant and anti-inflammatory functions31, found to be absent throughout erythropoiesis, in parallel to low gene expression levels (restricted to latest maturation stage) of the (constitutive) HO-2 isoform. These results are in line with previous observations from our group showing absence of the HO-1 protein on normal human BM erythroblasts31, despite the abundance of its strongest inducer (i.e., heme)32. Altogether, these results suggest differential heme-dependent regulation of HO-1 expression in erythroid cells, in order to prevent potentially redundant cycles of heme biosynthesis and degradation.
Because of significant production of pro-oxidant molecules during erythropoiesis such as heme and hemoglobin33 erythroblasts also require a tight regulation of the redox machinery. Such oxidative stress might contribute to progressively higher expression levels of genes involved in the metabolism of radical oxygen species (ROS), such as the UCP2 and PARK7 oxidative stress response genes. Other genes, found to be expressed throughout erythropoiesis, that are also involved in protecting cells from oxidative stress, included the PRDX3, TXNIP, DUSP1, GAB1, and GPX1 genes, which code for relevant antioxidant enzymes that act on hydrogen peroxide detoxification and protection of hemoglobin and other oxidative breakdown-associated proteins34,35.
Interestingly, an overall similar GEP was observed for both apoptosis and cell differentiation-associated genes suggesting a direct link between them. Thus, several genes within both functional groups were expressed in stage 1 precursors and decreased thereafter (e.g., KAT2A and PML), while other genes (e.g., PTPN6 and CYBRD1) were upregulated at the latest maturation stage, including the EPB42 (Erythrocyte Membrane Protein Band 4.2) gene involved in mechanical regulation and shaping of erythrocytes36, which progressively increased its expression from stage 2 to stage 3 NRBC. Apoptosis is a physiologic process essential for normal tissue development and differentiation37, which is typically associated with caspase activation, collapse of the nuclear envelope, DNA break-up into small fragments, and changes in cell shape and surface, leading to the exposure of signaling molecules to phagocytic cells37. Several caspases (e.g., CASP-3, CASP-8, CASP-9) have been found to be expressed during in vitro erythroblast differentiation38. Here we found continuous and stable expression of CASP-2 and putative expression of CASP-8 at the latest stage of maturation of NRBC precursors, in the absence of both CASP-3 and CASP-9; these observations suggest the involvement of death receptors instead of mitochondrial signaling pathways39 in erythropoietic apoptotic signaling cascades. Other apoptosis-associated genes, such as the TP53INP1 gene coding for the tp53inp1 tumor suppressor protein involved in the regulation of autophagy40, a critical process in erythropoiesis enucleation and removal of mitochondrial and other organelles41, were also expressed on both early and late (erythroid) maturation stages. In turn, the ATG2, ATG4, ATG9, mTOR, ATG1 (ULK1), and RB1CC1 genes, which are critical for mitochondrial and ribosomal clearance (e.g., mTOR and its downstream target ULK1 gene) and for autophagosome fusion (e.g., ATG4)42 and complex formation (e.g., ULK1 and RB1CC1) during terminal erythroid differentiation43, were progressively upregulated during erythroid maturation. Of note, expression of ULK1 was inversely related to that of PML, the later gene acting as an inhibitor of mTOR44. Other genes involved in the enucleation of erythrocytes such as the ID2 and RB1 genes, showed an expression profile restricted to the latest stage of maturation, in parallel to genes associated with oxygen homeostasis such as the EGLN1 and SOD2 genes.
A relative unexplored aspect of erythropoiesis is its relationship with immune response-associated genes. Several studies have highlighted the potential role of genes associated with the immune system, like FOXO3, in modulating the production and differentiation of erythroid precursors45. In addition to upregulation of FOXO3 at the more mature NRBC maturation stages, in line with previous observations46, functional analysis of such maturation stage-associated GEPs showed upregulation of several cell differentiation and immune response-associated genes at intermediate and late stages of maturation. Among other genes, these included genes that are triggered by (i) environmental stress conditions, leading to an increased cell permeability (e.g., BPI) and adhesion (e.g., CAECAM8) and (ii) the innate host defense (CRISP3)47,48,49. In turn, expression of other genes involved in response to proinflammatory cytokines and in the activation of the IkappaB kinase (IKK) within the NF-kappaB pathway, such as the TRAF6 gene, was restricted to the more mature stage 3 erythroid precursors; these findings are in contrast with previous (human) in vitro findings showing downregulation of TRAF6 during erythroid maturation10. In parallel to TRAF6, the DDIT3 transcription factor, involved in pro-apoptotic signaling associated to cell stress responses, was also upregulated in the more mature (stage 3) NRBC both here and in previous in vitro studies10.
Regarding the STAT gene family, slightly different GEP vs. previously in vitro data were found here10,50. Thus, our results showed a transiently lower STAT1 and STAT3 expression at stage 2 NRBC which recovered later on, while in vitro studies showed a continuous progressively lower expression of these genes10. Similarly, while STAT2 expression remained stable here, it appeared to be downregulated in previous in vitro studies10. In contrast, similar GEP were found here and in previous in vitro studies for the STAT5A (gene expression levels progressively downregulated along erythroid differentiation, being absent in the more mature erythroid precursors), and STAT5B (expressed at similar levels throughout the whole erythropoiesis). These differences observed between our ex vivo and previous in vitro studies in the expression pattern of STAT5 gene might be due to the different in vitro and in vivo activation conditions (e.g., erythropoietin doses). In line with this hypothesis, distinct ex vivo (absence of expression) and in vitro (upregulation) GEP10 were also observed for the GDF15 gene, a gene whose hyperexpression has been associated with both ineffective erythropoiesis or stressed erythropoiesis51. Altogether, these findings further support the relevance and complementarity of both types of in vitro and ex vivo studies.
In summary, here we provide for the first time detailed information about the ex vivo GEP of normal BM human erythroid precursor cells at different maturation stages. The results presented extend on and complement previous in vitro data, contributing to a better understanding of erythroid maturation and providing a frame of reference for the study of erythropoiesis in both neoplastic and non-neoplastic disease conditions.
Methods
BM samples
BM aspirated samples were obtained from healthy volunteers (2 males and 3 females; age range: 11–53 years). Prior to entering the study, each individual gave his written informed consent to participate according to the Declaration of Helsinki. The study was approved by the local ethics committee of the Cancer Research Centre (Salamanca, Spain).
Isolation of distinct subsets of normal human BM erythroid precursors
Fresh human BM samples were incubated with an ammonium chloride buffer to lyse mature non-nucleated red cells, washed twice in phosphate buffered saline (PBS), and stained with an 8-color combination of monoclonal antibodies, strictly following the EuroFlow standard operating procedures (SOPs)14. To stain BM cells, the following antibody reagents were used: CD34 (clone 8G12) conjugated with peridinin chlorophyll protein/cyanine 5.5 (Becton Dickinson Biosciences (BD), San Jose, CA), CD45 (clone HI30) conjugated with pacific orange (ExBio, Vestec, Czech Republic), CD36 (clone SLB-IVC7) conjugated with fluorescein isothiocyanate (BD), CD105 (clone 1G2) conjugated with phycoerythrin (Beckman Coulter (BC), Brea, CA), HLADR (clone L243) conjugated with pacific blue (BioLegend, San Diego, CA), CD117 (clone 104D2D1) conjugated with phycoerythrin/cyanine 7 (BC), CD33 (clone P67,6) conjugated with allophycocyanine (BD), and CD71 (clone M-A712) conjugated with allophycocyanine/cyanine 7 (BD). Stained cells were subsequently centrifuged (250 × g for 10 min), the cell pellet washed once (250 × g for 10 min) in 2 mL of PBS and resuspended in 0.5 mL of PBS. Stained samples were measured in either a FACSAria I or a FACSAria III flow cytometers (BD) equipped with the FACSDiva software (BD). Prior to sorting, 105 cells per BM sample were measured in the flow cytometer and used to define the gating regions (Fig. 1) for subsequent isolation of the distinct maturation-associated subsets of BM erythroid precursors. Overall, three distinct subpopulations of nucleated red blood cell precursors (NRBC) were identified and FACS-sorted, including (from the less differentiated to the more mature cells): (1) CD34+ HLADR+ CD117low CD36low CD45low CD105+ CD33− CD71low (stage 1) erythroid precursors (5.7% ± 0.5% of all NRBC); (2) CD34− HLADR− CD117− CD36+ CD45− CD105high CD33− CD71+ (stage 2) NRBC (9.6% ± 0.8% of all NRBC); and (3) CD34− HLADR− CD117− CD36high CD45− CD105+/− CD33− CD71high (stage 3) red cell precursors (84.7% ± 1.3% of all NRBC). For data analysis, the Infinicyt software program (Cytognos SL, Salamanca, Spain) was used. For each cell population, between 5600 and 500,000 cells were sorted directly into PBS, immediately centrifugated, and resuspended in 100 μL of RA1 buffer (NucleoSpin® RNA XS kit, Macherey-Nagel, Düren, Germany). The purity of the isolated cell populations (> 97%) was confirmed through re-analysis of ≈ 500 sorted cells per cell fraction and sample.
RNA extraction and gene expression profiling studies
GEPs were assessed on identical amounts of total mRNA extracted from each of the three purified subpopulations of BM NRBC. For this purpose, total mRNA was extracted using the NucleoSpin RNA XS kit, strictly following the instructions of the manufacturer. The quality and concentration of the extracted mRNA were evaluated using the Agilent 2100 Bioanalyzer and the RNA6000 Nanochips and Picochips (Agilent Technologies, Santa Clara, CA). GEPs were determined using the Affymetrix GeneChip Human Gene 1.0 ST Array (Affymetrix), strictly following the procedures recommended by the manufacturer.
For GEP data analysis, raw data (Affymetrix.CEL files) was normalized using the robust multi-array average (RMA) algorithm, which included background correction, quantile normalization, log2 transformation, and probe set summarization of GEP raw data. The Affymetrix GeneChip Human Gene 1.0 ST Array negative (n = 1195) and positive (n = 2904) control probes were measured in parallel; cut-off values for positive gene expression signals were set at the 1st lower quartile of the positive control gene probe signals (Supplementary Figure 1).
Genes that were differentially expressed between distinct maturation-associated subsets of BM erythroid precursors were ranked based on the comparison between their expression values in two consecutive stages of (erythroid) maturation, using the “Gene Set Enrichment Analysis” (GSEA) software52. Subsequently, the GeneCodis (Gene Annotation Co-occurrence Discovery; Madrid, Spain)53 annotation tool was used for gene ontology and functional enrichment analyses for those genes that were found to be expressed either at stable or variable levels during human BM erythroid maturation (Supplementary Figure 2).
Quality control of GEP of purified BM nucleated erythroid precursors at different stages of maturation
As an internal control for the GEP results obtained for each of the distinct cell populations, we evaluated the expression of genes that are known to be either specific or characteristic of erythroid cells vs. other hematopoietic cell lineages. Thus, hemoglobin genes (i.e., HBA, HBB, HBD, HBM, HBG, and HBQ1) were strongly expressed in the three distinct populations of maturation-associated erythroid precursors, while those genes coding for the embryonic hemoglobin chains (HBE1 and HBZ) were absent. In contrast, genes related with the neutrophil (e.g., CD16a or FcγRIIIa), lymphocyte (e.g., CD19 B-lymphocyte antigen), and monocytic (e.g., CD14 or myeloid cell-specific leucine-rich glycoprotein) cell lineages, were systematically absent in all different FACS-sorted maturation-associated subsets of BM erythroid precursors analyzed (Supplementary Figure 3). In order to validate microarray data, gene expression for the HMOX1, HMOX2, GYPA, ALAS1, and ALAS2 genes was investigated in parallel for all purified maturation stage-associated nucleated red cell populations by real-time PCR (ABI 7500 Real-Time PCR System, Applied Biosystems) using the SYBR Green I double-stranded DNA-specific fluorophore (Power SYBR® Green PCR Master Mix; Applied Biosystems) (Supplementary Figure 4). cDNAs were amplified using 40 cycles of PCR (denaturation for 15″ at 95 °C; annealing and extension for 60″ at 60 °C). Real-time PCR results were normalized by GAPDH mRNA expression levels, and Ct values were calculated using 2−ΔΔCt, where ΔΔCt = ΔCt “treatment” − ΔCt control.
References
Méndez-Ferrer, S. et al. Mesenchymal and haematopoietic stem cells form a unique bone marrow niche. Nature 466, 829–834 (2010).
Ney, P. A. Gene expression during terminal erythroid differentiation. Curr. Opin. Hematol. 13, 203–208 (2006).
Lu, S. J. et al. Biologic properties and enucleation of red blood cells from human embryonic stem cells. Blood 112, 4475–4484 (2008).
Palis, J. Primitive and definitive erythropoiesis in mammals. Front. Physiol. 5, 3 (2014).
Lapidot, T. & Petit, I. Current understanding of stem cell mobilization: the roles of chemokines, proteolytic enzymes, adhesion molecules, cytokines, and stromal cells. Exp. Hematol. 30, 973–981 (2002).
Hanspal, M. et al. Asynchronous synthesis of membrane skeletal proteins during terminal maturation of murine erythroblasts. Blood 80, 530–539 (1992).
Liang, R. & Ghaffari, S. Advances in understanding the mechanisms of erythropoiesis in homeostasis and disease. Br. J. Haematol. 174, 661–673 (2016).
Vaisman, B. et al. Ferritin expression in maturing normal human erythroid precursors. Br. J. Haematol. 110, 394–401 (2000).
Quigley, J. G. et al. Identification of a human heme exporter that is essential for erythropoiesis. Cell 118, 757–766 (2004).
An, X. et al. Global transcriptome analyses of human and murine terminal erythroid differentiation. Blood 123, 3466–3477 (2014).
Li, J. et al. Isolation and transcriptome analyses of human erythroid progenitors: BFU-E and CFU-E. Blood 124, 3636–3645 (2014).
Machherndl-Spandl, S. et al. Molecular pathways of early CD105-positive erythroid cells as compared with CD34-positive common precursor cells by flow cytometric cell-sorting and gene expression profiling. Blood Cancer J. 3, e100 (2013).
Kolbus, A. et al. Cooperative signaling between cytokine receptors and the glucocorticoid receptor in the expansion of erythroid progenitors: molecular analysis by expression profiling. Blood 102, 3136–3146 (2003).
van Dongen, J. J. et al. EuroFlow antibody panels for standardized ndimensional flow cytometric immunophenotyping of normal, reactive and malignant leukocytes. Leukemia 26, 1908–1975 (2012).
McGrath, K. E., Catherman, S. C. & Palis, J. Delineating stages of erythropoiesis using imaging flow cytometry. Methods 112, 68–74 (2017).
Ferreira, R., Ohneda, K., Yamamoto, M. & Philipsen, S. GATA1 function, a paradigm for transcription factors in hematopoiesis. Mol. Cell. Biol. 25, 1215–1227 (2005).
Witthuhn, B. A. et al. JAK2 associates with the erythropoietin receptor and is tyrosine phosphorylated and activated following stimulation with erythropoietin. Cell 74, 227–236 (1993).
Elliott, S., Pham, E. & Macdougall, I. C. Erythropoietins: a common mechanism of action. Exp. Hematol. 36, 1573–1584 (2008).
Kuhrt, D. & Wojchowski, D. M. Emerging EPO and EPO receptor regulators and signal transducers. Blood 125, 3536–3541 (2015).
Zhang, J., Socolovsky, M., Gross, A. W. & Lodish, H. F. Role of Ras signaling in erythroid differentiation of mouse fetal liver cells: functional analysis by a flow cytometry-based novel culture system. Blood 102, 3938–3946 (2003).
Rogers, H. M. et al. Hypoxia alters progression of the erythroid program. Exp. Hematol. 36, 17–27 (2008).
Akashi, K. et al. Transcriptional accessibility for genes of multiple tissues and hematopoietic lineages is hierarchically controlled during early hematopoiesis. Blood 101, 383–389 (2003).
Cross, M. A. & Enver, T. The lineage commitment of haemopoietic progenitor cells. Curr. Opin. Genet. Dev. 7, 609–613 (1997).
Rylski, M. et al. GATA-1-mediated proliferation arrest during erythroid maturation. Mol. Cell. Biol. 23, 5031–5042 (2003).
Rodriguez, P. et al. GATA-1 forms distinct activating and repressive complexes in erythroid cells. EMBO J. 24, 2354–2366 (2005).
Gregory, T. et al. GATA-1 and erythropoietin cooperate to promote erythroid cell survival by regulating bcl-xL expression. Blood 94, 87–96 (1999).
Welch, J. J. et al. Global regulation of erythroid gene expression by transcription factor GATA-1. Blood 104, 3136–3147 (2004).
Gautier, E. F. et al. Comprehensive proteomic analysis of human erythropoiesis. Cell Rep. 16, 1470–1484 (2016).
Cox, T. C., Bawden, M. J., Martin, A. & May, B. K. Human erythroid 5-aminolevulinate synthase: promoter analysis and identification of an iron-responsive element in the mRNA. EMBO J. 10, 1891–1902 (1991).
Chao, C. Y. et al. Induction of heme oxygenase-1 and inhibition of TPA-induced matrix metalloproteinase-9 expression by andrographolide in MCF-7 human breast cancer cells. Carcinogenesis 34, 1843–1851 (2013).
Alves, L. R. et al. Heme-oxygenases during erythropoiesis in K562 and human bone marrow cells. PLoS ONE 6, e21358 (2011).
Ponka, P. Tissue-specific regulation of iron metabolism and heme synthesis: distinct control mechanisms in erythroid cells. Blood 89, 1–25 (1997).
Hattangadi, S. M. & Lodish, H. F. Regulation of erythrocyte lifespan: do reactive oxygen species set the clock? J. Clin. Invest. 117, 2075–2077 (2007).
Liu, Y. X. et al. DUSP1 is controlled by p53 during the cellular response to oxidative stress. Mol. Cancer Res. 6, 624–633 (2008).
He, X. & Ma, Q. Redox regulation by nuclear factor erythroid 2-related factor 2: gatekeeping for the basal and diabetes-induced expression of thioredoxin-interacting protein. Mol. Pharmacol. 82, 887–897 (2012).
Das, A. K. et al. Human erythrocyte membrane protein 4.2 is palmitoylated. Eur. J. Biochem. 224, 575–580 (1994).
Elmore, S. Apoptosis: a review of programmed cell death. Toxicol. Pathol. 35, 495–516 (2007).
Zermati, Y. et al. Caspase activation is required for terminal erythroid differentiation. J. Exp. Med. 193, 247–254 (2001).
Chinnaiyan, A. M. The apoptosome: heart and soul of the cell death machine. Neoplasia 1, 5–15 (1999).
Seillier, M. et al. TP53INP1, a tumor suppressor, interacts with LC3 and ATG8-family proteins through the LC3-interacting region (LIR) and promotes autophagy-dependent cell death. Cell Death Differ. 19, 1525–1535 (2012).
Zhang, J. et al. Autophagy as a regulatory component of erythropoiesis. Int. J. Mol. Sci. 16, 4083–4094 (2015).
Betin, V. M. et al. Autophagy facilitates organelle clearance during differentiation of human erythroblasts: evidence for a role for ATG4 paralogs during autophagosome maturation. Autophagy 9, 881–893 (2013).
Jung, C. H. et al. ULK-Atg13-FIP200 complexes mediate mTOR signaling to the autophagy machinery. Mol. Biol. Cell. 20, 1992–2003 (2009).
Bernardi, R. et al. PML inhibits HIF-1alpha translation and neoangiogenesis through repression of mTOR. Nature 442, 779–785 (2006).
Liang, R. et al. A systems approach identifies essential FOXO3 functions at key steps of terminal erythropoiesis. PLoS Genet. 11, e1005526 (2015).
Franco, S. S. et al. Resveratrol accelerates erythroid maturation by activation of FoxO3 and ameliorates anemia in beta-thalassemic mice. Haematologica 99, 267–275 (2014).
Beamer, L. J. Structure of human BPI (bactericidal/permeability-increasing protein) and implications for related proteins. Biochem. Soc. Trans. 31, 791–794 (2003).
Zhao, L. et al. Purification and characterization of a 95-kDa protein—carcinoembryonic antigen-related cell adhesion molecule 8—from normal human granulocytes. J. Immunol. Methods 270, 27–35 (2002).
Udby, L. et al. An ELISA for SGP28/CRISP-3, a cysteine-rich secretory protein in human neutrophils, plasma, and exocrine secretions. J. Immunol. Methods 263, 43–55 (2002).
Cokic, V. P. et al. JAK-STAT and AKT pathway-coupled genes in erythroid progenitor cells through ontogeny. J. Transl. Med. 10, 116 (2012).
Kim, A. & Nemeth, E. New insights into iron regulation and erythropoiesis. Curr. Opin. Hematol. 22, 199–205 (2015).
Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102, 15545–15550 (2005).
Nogales-Cadenas, R. et al. GeneCodis: interpreting gene lists through enrichment analysis and integration of diverse biological information. Nucleic Acids Res. 37, W317–W322 (2009).
Acknowledgements
The authors would like to thank Prof. Iqbal Hamza from Maryland University for his valuable comments; all technicians from the Cytometry Service, Institute of Paediatrics and Puericulture Martagão Gesteira (IPPMG), UFRJ, Rio de Janeiro, Brazil and Carlos Pietro from Cancer Research Centre (IBMCC, USAL-CSIC), Institute for Biomedical Research of Salamanca (IBSAL), USAL, Salamanca, Spain. This work was supported by the following grants: Bilateral Cooperation Program between Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES (Brasília/Brazil) and Dirección General de Políticas Universitárias – Ministério de Educación, Cultura y Deportes - DPGU (Madrid/Spain) (311/15) and RD12/0036/0048 from RETICS (Instituto de Salud Carlos III, Ministerio de Economia y Competitividad, Madrid, Spain and Fondos FEDER); Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro of Brazil (E26/110.105/2014; E26/102.191/2013); Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPQ of Brazil (400194/2014-7). F.V.M. was supported by grant from CAPES (Brazil). Funders had no role in the design of the study, data collection and analysis, decision to publishing nor in the preparation of the manuscript.
Author information
Authors and Affiliations
Contributions
F.V.M., E.S.C., L.R.A., and A.O. conceived and designed the experiments. F.V.M., M.L.S., and P.B. performed the flow cytometry cell sorting experiments. F.V.M. and C.T. performed all others experiments. F.V.M., R.T.P., C.E.P., and M.G.P.L. analyzed the data. F.V.M., L.R.A., E.S.C., and A.O. wrote the paper.
Corresponding authors
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Edited by N. Barlev
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Mello, F.V., Land, M.G.P., Costa, E.S. et al. Maturation-associated gene expression profiles during normal human bone marrow erythropoiesis. Cell Death Discov. 5, 69 (2019). https://doi.org/10.1038/s41420-019-0151-0
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41420-019-0151-0
- Springer Nature Limited
This article is cited by
-
Recent updates of stem cell-based erythropoiesis
Human Cell (2023)
-
Phiclust: a clusterability measure for single-cell transcriptomics reveals phenotypic subpopulations
Genome Biology (2022)
-
Erythroid lineage-specific lentiviral RNAi vectors suitable for molecular functional studies and therapeutic applications
Scientific Reports (2022)
-
Heme metabolism genes Downregulated in COPD Cachexia
Respiratory Research (2020)